die junejuly 2014

E U R O P E

JULY 2014

D i a g n o s i s • T e c h n o l o g y • T h e r a p y • p r e v e n T i o n

radiology education Fusing MRI with Ultrasound images— a tool to enhance ultrasound education

Ultra high Field MriEnsuring a continuing supply of magnets for ultra high-field 7 Tesla MRI systems

personalizing an integrated approach to cT imaging

The role of optimized Molecular imaging in personalized cancer Medicine

hospital FocusA vibrant health care center in Eastern Belgium

industry news

Technology update

UltrasoUnd Elastography in scrEEning and diagnosis of BrEast cancEr

Contrast-enhanCed speCtral mammography: a promising breast imaging tool

DI EUROPE

EXPERIENCE A TRULY PREMIERULTRASOUND SYSTEMRS80A, a radiology ultrasound, increases diagnostic confidence utilizing advanced technologies developed by Samsung,while simultaneously enhancing user experience and providing an efficient work environment.

BY AlAn BARclAY, Ph.D.

FROM THE EDITOR

The paradox is simple to describe but much more difficult to resolve. To avoid being pigeon-holed as a mere back-room sup-

plier of reports describing cases presented on screens far from the patient, the modern radiologist must actively undertake more and more other activities in addition to the primary responsibility of interpreting all imaging studies presented to him in a timely manner. Increasingly radiologists are hav-ing to shoulder multiple additional respon-sibilities such as examination protocoling, supervising contrast agent and nuclear medicine injections, ultrasound scanning, clinical consults, participating in multi-dis-ciplinary meetings and the communication of imaging findings to referring clinicians. Very few radiologists would argue against the necessity of such non-interpretive activ-ities, and indeed most welcome them as a recognition of a fuller, more satisfying and rewarding role than simply dictating report after report. The problem is finding the time to do all this. As radiology becomes more and more an unavoidable hospital stage through which almost all patients must and do pass, the pressure on the radi-ologist to churn out reports becomes more intense. Whereas in many other sectors an ever-increasing work-load is often con-sidered as a welcome, positive sign of the vitality of the profession (not to mention an increasing source of revenue) in radiology increasing work-loads are often considered just as an increasingly difficult burden to be shouldered and to be managed as efficiently as possible. Of course, the one radical solution to this problem, namely a drastic reduction in the basic demand — or obsession — for imaging studies is at best a forlorn hope in this world where it is important to share risks. A reduction in the demand for images can also be seen as some sort of sacrilegious insult to the primordial role of radiology. So, in reality, the conflict-ing demands of being called upon for more and more non-interpretive activities while at the same time being required to gener-ate many reports are resolved in a purely empirical way. This involves simply trying as much as possible to fit everything in and

inevitably involves regular interruptions in the basic work-flow. While, with their experience of innumerable on-call sessions during their education and post-graduate training, most clinicians regard interrup-tions as par for the course, the fact is that constant interruptions are a well-recognized source of inefficiencies and, much worse, as a source of errors. The literature in behav-ioural science is replete with the — mostly negative — studies on the effects of inter-ruptions on people carrying out complex cognitive tasks. Often this can entail for the “interruptee”, a sense of having to recover afterward. An additional irony is that, natu-rally enough, the probability of being inter-rupted at least once while performing a task increases with the length of time required to complete the task. In radiological terms the reading of a complex study which demands more continual and focussed attention is this more likely be interrupted than simple or routine tasks. Only now are studies appearing on the specific analysis of the effect of constant interruptions on radi-ologists, such as in a recent report by Yu and co-workers on their analysis of the number and nature of interruptions to radi-ologists in their clinic. (Yu JP, Kansagra AP, Mongan J The Radiologist’s Workflow Environment: Evaluation of Disruptors and Potential Implications J Am Coll Radiol. 2014 Jun;11(6):589-93). Yu et al. recognized that, although communicating with refer-ring providers is perhaps one of the most important aspects of radiologists’ responsi-bilities, it is also, unfortunately, one of the most disruptive. While there is a vicarious fascination of seeing the actual quantitation of the numbers of interruptions that Yu et al. report in their hospital, which of course varies hugely from institution to institution, the key question to be addressed by such studies is whether, apart from the general fatigue and job dissatisfaction of the radi-ologist, the interruptions actually affect the quality of the final report in objective terms such as diagnostic accuracy. Here unsur-prisingly, there is no quantitative conclusion at least yet. Intuitively however there is a gnawing feeling that constant interruptions can only be detrimental to quality. n

VOlUME 30, nUMBER 4

eDiTorial aDvisory BoarD

Andreas Adam, London Richard P. Baum, Bad Berka Frits h. Barneveld Binkhuysen, Elias Brountzos, Athens Amersfoort Filipe caseiro Alves, Coimbra carlo catalano, Rome Maksim cela, Tirana Patrick cozzone, MarseilleKatarzyna Gruszczynska, Anne Grethe Jurik, Arhus KatowiceAndrea Klauser, Innsbruck Gabriel Krestin, RotterdamGabriele Krombach, Giessen christiane Kuhl, BonnPhilippe lefere, Roeselare heinz U. lemke, Kuessabergluis Martí-Bonmatí, Valencia Thoralf niendorf, Berlinchristiane nyhsen, Sunderland Anne Paterson, BelfastAnders Persson, Linköping hans Ringertz, Stockholm Gustav von Schulthess, Zurich Valentin E. Sinitsyn, Moscow Patrick Veit-haibach, Lucerne Thomas J. Vogl, Frankfurt

eDiTorial sTaFF

Editor Alan Barclay, Ph.D. US consulting Editor Greg Freiherr

Editorial coordinator Marika Cooper

Publisher David lansdowne Associate Publisher Bob Warren

eDiTorial conTacT 1421 Ophain, BelgiumTel. +32 479 370 364E-mail: [email protected]: [email protected]

sUBscriBer servicesTelephone: +44 1442 877777; Fax: +44 1442 870617

E-mail: [email protected]

inTernaTional sales oFFices

europe, north america & JapanDI Europe Ltd


Telephone: +(44) 1442 877777

Fax: +(44) 1442 870617

Contact: David Lansdowne

Contact: Bob Warren


Contact: Marika Cooper


china KoreaAdept Marketing Young Media Inc

Unit B, 13/f, Por Yen Building, 407 Jinyang Sangga

No 478 Castle Peak Road, 120-3 Chungmuro 4 ga

Cheung Sha Wan, Kowloon, Hong Kong Chung-Ku, Seoul, Korea 100-863

E-mail: [email protected] E-mail: [email protected]

Telephone: +852 2891 7117 Telephone: +82 2 2273 4819

Fax: +852 2893 2101 Fax: +82 2 2273 4866

Contact: Adonis Mak Contact: Young J. Baek

Article Reprint Sales DI Europe Ltd Tel: +44 1442 877777

Email: [email protected]

DIAGnOSTIc IMAGInG EUROPE is published eight times a year by DI Europe ltd Printed by Manson, St-Albans, UK. Annual subscriptions are available for €60 within Europe where it is also sent free of charge to physicians and radiology department heads. Outside of Europe, there is an annual subscription charge of €110 for air mail. Single copy price is €10. Editorial Advisory Board members suggest topics for coverage and answer questions for the editors. They do not conduct a formal peer-review of all manuscripts submitted to DI Europe.

copyright © 2014 DI Europe ltd. All rights reserved. Reproduction in any form is forbidden without express permission of copyright owner.

DI Europe ltd2 claridge courtlower Kings RoadBerkhamsted, herts hP4 2AFUKTelephone: +44 1442 877777Fax: +44 1442 70617

The increasing rigors of the typical radiology workload: the effects of constant interruptions

JULY 2014 D I E U R O P E 3

DI EUROPE

DIEUROPE.cOm

EXPERIENCE A TRULY PREMIERULTRASOUND SYSTEMRS80A, a radiology ultrasound, increases diagnostic confidence utilizing advanced technologies developed by Samsung,while simultaneously enhancing user experience and providing an efficient work environment.

COMING SOON IN The NexT SepT/OCT ISSue: health IT MRInuclear Medecine Women healthJFR Preview

COVeR STORY

contrast-EnhancEd spEctral mammography: a promising BrEast imaging toolA recent study in women referred from a breast cancer screening program showed that cESM is a powerful tool enabling false-positive findings to be ruled out quickly, and true-positive cases to be detected with a high degree of confidence.

By Dr M Lobbes � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � � 20

IMAGING NeWS

huge study confirms efficacy of tomography plus digital mammography in breast screening. . . . . . . . . . . . . . . . . . . . . . . . . . . 5lifetime cancer risk from heart imaging is low for most children, but more complex tests raise risk. . . . . . . . . . . . . . . . . . . . . . . . . . . . 5PSMA-based imaging identifies even treatment-resistant prostate cancer . . . . 7PET/MRI improves diagnostic certainty in breast cancer . . . . . . . . . . . 7Iterative reconstruction techniques reduce radiation dose for pediatric brain cT . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8A new diagnostic tool for dementia diseases. . . . . . . . . . . . . . . . . . . . 8Molecular imaging progress in RA research . . . . . . . . . . . . . . . . . . . . 8SPEcT/cT of sentinel lymph nodes more sensitive than lymphoscintigraphy . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9MR guided focused ultrasound reduces cancer pain . . . . . . . . . . . . . . 9Resting state fMRI brain scans detect early Parkinson’s Disease. . . . 10

JuLY 2014

®

E U RO P E

RepORTS

fusing mri with Ultrasound images— a tool to enhance ultrasound education and improve clinical ultrasound

page 14

Ensuring a continuing supply of magnets for ultra high-field 7 tesla mri systems

page 12

ReGuLARS

3| FROM The eDITOR

16| peRSONALIzING AN INTe-GRATeD AppROACh TO CT IMAGING

28| The ROLe OF OpTIMIzeD MOLeCuLAR IMAGING IN peRSONALIzeD CANCeR MeDICINe

32| hOSpITAL FOCuS

A vibrant health care center in Eastern BelgiumThe Ziekenhuis Oost-limburg (ZOl) hospital is experiencing a steady increase in patients and, in the radiology department, there is a corresponding rise in the number of examinations being carried out.

41| INDuSTRY NeWS

44| TeChNOLOGY upDATe

Ultrasound Elastography in screening and diagnosis of Breast cancer

page 36


DI EUROPEVISIT US AT

DIEUROPE.cOm

NeWSIMAGING


huge study confirms efficacy of tomography plus digital mammography in breast screening

lifetime cancer risk from heart imaging is low for most children, but more complex tests raise risk

3D Mammography finds significantly more invasive cancers and reduces unnecessary recalls, according to a large, retrospective study published in the June 25 issue of the JAMA (Friedewald et al Breast Cancer Screening Using Tomosyn-thesis in Combination With Digital Mam-mography. JAMA. 2014;311(24):2499) The study, the largest of its kind, focused on the impact of 3D mammography at a diverse range of sites across the U.S, looking at nearly half a million mam-mograms at 13 sites. Key findings were a 41% increase in invasive cancer detected with 3D mammography; a 15% decrease in unnecessary recalls for false positives; and a 29% increase in the detection of all breast cancers.

“This study confirms what we already know: 3D mammography finds more of the invasive, harmful cancers we want to find and saves women the anxiety and cost of having additional exams for what turns out to be a false alarm,” said the study’s co-author Donna Plecha, MD. “We already knew that breast screening saves lives and this study provides us with firm data that 3D mammog-raphy is a better test for detecting breast cancer early when it is treat-able.” Hologic’s 3D mammography (breast tomosynthesis) system was used exclusively in the study, as this is the only FDA-approved 3D mam-mography. The system combines advanced digital mammography and tomosynthesis-generated images to provide a more detailed, highly focused picture of the breast. These images are then used to produce a series of one-millimeter thick slices that can be viewed as a 3D recon-struction of the breast. Women see little difference between a conven-tional 2D mammogram and a 3D mammogram. The exam takes just a few seconds longer and the position-ing is the same. The technology gives

radiologists the ability to identify and characterize individual breast struc-tures and clearly see features which might be obscured in a traditional 2D mammogram by overlapping normal breast anatomy that may mimic or mask a tumor. Dense tissue and over-lapping tissue structures may lead to false positive or false negative results with standard mammography.

Breast cancer remains a significant health problem and statistics indicate that one in eight women will develop the disease in her lifetime. The stage at which the cancer is discovered influ-ences a woman’s chance of survival and annual mammography after the age of 40 enables physicians to identify even the smallest abnormalities. In fact, when breast cancer is detected early and confined to the breast, the five-year survival rate is 97 percent. “Breast cancers caught in the initial stages by mammography are more likely to be cured and are less likely to require chemotherapy or as extensive surgery,” said Dr. Plecha. “This study shows that 3D mammography is a more effective screening tool, and we must make it accessible to all women.”http://tinyurl.com/Friedewald-et-al-paper

Radiation from standard X-rays is rel-atively low and doesn’t significantly raise lifetime cancer risks for most young chil-dren, according to a recent paper (John-son et al. Cumulative Radiation Exposure and Cancer Risk Estimation in Children with Heart Disease. Circulation 2014 Jun 9.113.005425) from Duke University Medical Center in Durham, NC USA. The researchers followed 337 children under age 6 who had had surgery for heart disease. Their operations required almost 14,000 imaging procedures, including X-rays, CT scans, and cardiac catheterization using fluoroscopy.

The study is the first in which research-ers quantified cumulative radiation doses in pediatric heart patients and predicted lifetime cancer risks based on the type of exposure. The researchers found that for the average child in the study, the cumulative effective dose of ioniz-ing radiation was relatively low — less than the annual background radiation exposure. However, some children with complex heart disease can be exposed to large cumulative doses that increase the estimated lifetime risk of cancer up to 6.5 percent above baseline.“There are definitely times when radiation is neces-sary,” said Dr Hill study lead author “But it’s important for parents to ask and com-pare in case they can avert potentially high exposure procedures. Often, there are alternative or modified procedures

IMAGING NeWS


with less radiation, or imaging itself may not actually be necessary.”

In the study, the researchers reviewed medical records to find the most com-mon imaging procedures, calculated how much radiation the organs absorbed during each procedure, then used a US National Academy of Sciences report to analyze lifetime cancer risks based on the amounts of exposure. The study results showed that although X-rays accounted for 92 percent of imaging exams, 81 percent of the overall radiation exposure was a result of cardiac catheterizations and CT scans. Overall, lifetime cancer risk was found to increase from 0.002 percent for chest X-rays to 0.4 percent for complex CT scans and cardiac cath-eterization. Because they are more prone to breast and thyroid cancers, girls have double the cancer risks of boys. http://tinyurl.com/Johnson-et-al-Paper

psMa-based imaging identifies even treat-ment-resistant pros-tate cancer

Anti-androgen hormonal therapy, also known as chemical castration, can be an important defence against further disease progression for patients with prostate cancer that has metastasized. However some cases simply do not respond to the treatment. In an effort to improve clinical decision making, a ground-breaking molecular imaging agent has been developed to identify whether prostate cancer is responding favorably or not to therapy, according to work presented by a group from Johns Hopkins University at the recent US Society of Nuclear Medicine and Molec-ular Imaging’s 2014 Annual Meeting.

The patients were examined by PET/CT imaging using a F-18 labelled ligand known as DCFBC. This is a unique small-molecule that binds to prostate-specific membrane antigen (PSMA), which is more expressed on malignant prostate cells than on normal cells. The study showed the effectiveness of the imaging agent by providing substantial clinical data for both castration-sensitive and castration-resistant prostate cancer patients.

“Currently there is a great unmet need in prostate cancer management and drug development for a functional imaging agent that is able to detect prostate cancer and monitor response to therapy,” said Dr S Cho from the department of radiol-ogy at Johns Hopkins. “Unfortunately, a truly reliable functional imaging agent for prostate cancer does not exist. How-ever several exciting metastatic cancer imaging agents have been in develop-ment over the last several years. We are working toward improvements beyond the current capabilities of conventional bone and CT imaging, and a small-mole-cule PSMA-based PET radiopharmaceu-tical such as F-18 DCFBC is one such possibility.”

This study included the first 12 patients from an ongoing trial, includ-ing five cases of castration-sensitive and seven cases of castration-resistant can-cer, both with rising PSMA levels and evidence of metastases. Hot spots rep-resenting tumor uptake were correlated with serum prostate-specific antigen and folate levels as well as castration-resis-tant status. Results of the study showed F-18 DCFBC uptake was comparable to conventional imaging in relation to the lymph nodes, some bone and viscera, including the adrenal glands and pan-creas. Lower DCFBC uptake was seen in highly scarred bone metastases when compared to other kinds of growths, but DCFBC PET was found to be more sensitive than conventional imaging for detecting bone metastases, especially within the cervical spine and areas show-ing degenerative changes, as well as in subcentimeter-sized lymph nodes. Addi-tionally, a higher uptake of the agent was observed in castration-resistant bone metastases, and a direct link was found between PSMA levels and tumor-agent uptake. http://tinyurl.com/JHU-group-presentation

peT/Mri improves diagnostic certainty in breast cancer

Using the combined approach of PET and MRI, the key processes in the development of breast cancer can be visualized, allowing every sec-ond unnecessary breast biopsy to be avoided. This is the most significant finding of a study from the University Department of Radiology and Nuclear Medicine carried out in cooperation with other departments and centers at the MedUni Vienna (Pinker K, et al. Improved Differentiation of Benign and

Malignant Breast Tumors with Multi-parametric 18Fluorodeoxyglucose Posi-tron Emission Tomography Magnetic Resonance imaging: A Feasibility Study. Clinical Cancer Research 2014). Dr Pinker was able to demonstrate in the recently published clinical study — the first of its kind worldwide — that multi-parametric PET / MRI can achieve a diagnostic certainty of 96 per cent. This could mean that half of all breast biop-sies of benign nodes could be avoided . The diagnosis of breast cancer was made for the first time using a combi-nation of 3-Tesla MRI and FDG PET. By combining the two imaging methods, a variety of different information on the key processes involved in the develop-ment of breast cancer can be gleaned at the same time. The current study dem-onstrated that this multi-parametric PET/MRI allows a better non-invasive diagnosis of breast tumors: “It means we are able to distinguish more easily between benign and malignant tumors, thereby significantly reducing the num-ber of false positives.”www.meduniwien.ac.at

Higher diagnostic certainty of the PET/MRI system could mean that half of all current biopsies (of benign nodes) could be avoided

IMAGING NeWS

8 D I E U R O P E JULY 2014

iterative reconstruc-tion techniques reduce radiation dose for pediatric brain cT

A study conducted at Massachusetts General Hospital and Harvard Medical School found that estimated radiation doses are substantially lower for pedi-atric CT exams of the brain when an adaptive statistical iterative reconstruc-tion technique (ASIR) was used com-pared to those that did not use ASIR. The researchers found that the brain and salivary gland doses were much lower for ASIR-enabled exams compared to those without ASIR technique. However, no differences in the estimated organ doses were found for the thyroid gland, skeleton, and eye lenses across these two cohorts of CT exams. “CT radiation dose is an important concern with all imag-ing sites, especially for children,” said Dr RDA Khawaja. The results of this study were presented by Dr Khawaja at the recent annual meeting of the American Roentgen Ray Society (ARRS). “We per-formed this study to do a preliminary analysis of pediatric head CT examina-tions and to assess the factors influencing radiation doses.”

Mean radiation dose was 1.6 ± 1.5 mSv (estimated effective dose) in pedi-atric head CT. In addition to the iterative reconstruction algorithm, patient age and effective body diameter significantly affected the doses.

The results of this study were pre-sented by Dr Khawaja at the recent annual meeting of the American Ront-gen Ray Society (ARRS) www.arrs.org

a new diagnostic tool for dementia diseases

A new diagnostic tool helps clinicians to differentiate between Alzheimer’s dis-ease, frontotemporal dementia and mild cognitive impairment. Presented by Dr M. Ruiz at the University of Eastern Fin-land, the new method consists of a Dis-ease State Index combining data from multiple sources, and of a Disease State Fingerprint showing the findings in a visual format.

It is estimated that more than 35.6 million people are living with dementia worldwide. This number will increase in the coming years and there is a need to identify these patients to provide them with proper treatment as early as pos-sible. The differential diagnosis of the dementia diseases represents a challenge particularly in the early phases. Many studies have focused on predicting the possible conversion from mild cogni-tive impairment, a pre-dementia stage, to Alzheimer’s disease (AD), the most common dementia disease. Several methods have also been proposed for differentiating between AD and fronto-temporal dementia (FTD), another rela-tively common degenerative dementia. An early and precise diagnosis of these two dementia diseases is needed in order to benefit from treatments designed to influence the disease mechanisms.

In recent years, important advances have been made especially in the devel-opment of new diagnostic methods. Sev-eral biomarkers and tests are used in clin-ical practice, such as cerebrospinal fluid biomarkers, imaging methods, genetic profiling and neuropsychological tests. However, making a differential diagnosis is not easy due to overlapping clinical and biomarker findings in the interpre-tation of all this multitude of data. Fur-thermore, there is no single biomarker or test which could clearly define whether a patient is suffering from AD or FTD.

The approach of Dr Ruiz (Structural MRI in Frontotemporal Dementia: Com-parisons between Hippocampal volu-metry, Tensor-based morphometry and Voxel-based morphometry. PLoS ONE 7(12): e52531). introduces a new combi-nation of different methods for the dif-ferential diagnosis of AD, mild cognitive impairment and FTD, and describes a tool comprising a Disease State Index and its visual counterpart, a Disease State Fingerprint.

The software combines data from multiple sources such as psychological tests and brain MRI, and uses this data to create a Disease State Index, with a range of values between 0 and 1. In a healthy person, the index is close to 0, while an index close to 1 is an indicator of demen-tia. The Disease State Fingerprint shows the findings in an easy-to-interpret for-mat in which the key findings are clearly indicated by color and size.

With the help of the new diagnostic tool, clinicians could determine which methods are more relevant for profiling a patient with a certain dementia disease, e.g. whether it is mild cognitive impair-ment, FTD or AD. Even the first visit, the clinician could make a first diagnosis to initiate treatment and to counsel to the patient.http://tinyurl.com/Ruiz-Univ-E-Finland

Molecular imaging progress in ra research

Rheumatoid arthritis causes chronic pain for almost half of adults by the time they retire. It is estimated that one in five adults and almost 50 percent of those age 65 years or older have been clini-cally diagnosed with arthritis. However a new molecular imaging technique can visualize inflammation in the joints, giv-ing doctors a clear read on chronic pain and possible joint destruction, according to a recent report by a group of Dutch researchers at the Society of Nuclear Medicine and Molecular Imaging’s 2014 Annual Meeting.

In order to image arthritis inside the joints, researchers used multiple molecu-lar imaging systems, PET and SPECT, both of which image physiological pro-cesses. In this case researchers evaluated anti-fibroblast activation protein (FAP) antibodies involved in the inflammation

IMAGING NeWS


associated with rheumatoid arthritis. This was made possible with radiotracers that combine the molecular compound 28H1, which can bind to FAP in the body, with the radionuclides In-111, used in conjunction with SPECT imaging sys-tems, and with Zr-89, used with PET systems. “This research is novel because radiolabeled anti-FAP antibodies have never been used before in molecu-lar imaging for rheumatoid arthritis,” remarked Dr P Lavermann from the Department of Radiology and Nuclear Medicine at Radboud University Medical Center in Nijmegen, The Netherlands. “These antibodies are used for cancer imaging, but can also be used to image FAP expressed on activated fibroblasts in arthritic joints. We found a high accu-mulation of radiolabeled anti-FAP anti-bodies in arthritic joints using SPECT and PET imaging.” This was a preclini-cal study using small animal scanners. Results of the research showed that both In-111 28H1 and Zr-89 28H1 showed significantly increased imaging agent uptake in inflamed joints. In fact, that uptake was three to four times higher with these agents than another antibody agent evaluated as a control. Researchers also evaluated the commonly used imag-ing agent, FDG, to image the inflamma-tion, but uptake of this agent was not correlated with the severity of inflam-mation. This experimental model proved that 28H1 tagged with either In-111 or Zr-89 is a superior method for imaging arthritis.

“To the best of our knowledge, high-contrast images of this kind were unheard of until now,” said Laverman. He esti-mated that it may take two or more years to accumulate enough research to get the agents approved for arthritis imaging in mainstream clinical practice.http://tinyurl.com/SNNM-annual-meeting

specT/cT of sentinel lymph nodes more sen-sitive than lymphoscin-tigraphy

Startling data from an international multi-center trial and presented at the recent 2014 annual meeting of the Society of Nuclear Medicine and Molecular Imaging (SNMMI) provide growing evidence that sentinel node imaging is more effectively carried out with (SPECT/CT) hybrid functional imaging than with lymphoscintig-raphy. This conclusion is valid for a range of cancers displaying a variety of lymphatic drainage types associated with melanoma; breast carcinoma; and malignancies of the pelvis, such as prostate and cervical cancer.

Molecular imaging of sentinel lymph nodes can provide a surgi-cal map that can improve a patient’s chances of becoming cancer free. Lymph node imaging is important in the context of surgical resection, since cancerous cells spread first to the sen-tinel lymph nodes, before navigating the bloodstream and developing new malignancies elsewhere in the body. “We found significantly more sentinel lymph node involvement with SPECT/CT, which altered surgical planning for many of our patients—a finding that was repeated across all malignan-cies and clinical institutions,” said Dr T. Pascual, co-author of the study and a research scientist from the section of nuclear medicine and diagnostic imaging and division of human health of the International Atomic Energy Agency in Vienna, Austria. “These

results could potentially affect new clinical practice and shape appropriate use of SPECT/CT imaging for patients selected for surgery.”

The study findings showed that in breast cancer SPECT/CT imaging detected 13 percent more cancerous sentinel nodes—2,165 nodes com-pared with 1,892 using planar lym-phoscintigraphy. The hybrid SPECT system also detected 11.5 percent more sentinel nodes when imaging for melanoma. As for pelvic cancer, 29.2 percent more nodes were imaged using SPECT/CT than with planar imaging.

Changes in surgical planning as dictated by SPECT/CT results were substantial—16.9 percent of breast cancer surgeries underwent a change in management, 37 percent of surger-ies for melanoma changed and 64.1 percent of surgical plans for pelvic cancer were changed due to detec-tion of additional sentinel nodes. Cal-culated mismatch between sentinel nodes and lymphatic territories using the two imaging systems was gauged at 17 percent for breast carcinoma, 11.2 percent for melanoma and 50 percent for pelvic imaging. The significantly higher mismatch in pelvic tumors was thought to be due to relatively deeper lymphatic drainage and location of pelvic sentinel nodes.

Mr guided focused ultrasound reduces cancer pain

When cancer spreads to the bones, patients often suffer debili-tating pain. Now, the results of a newly published phase III clini-cal trial show that magnetic reso-nance guided focused ultrasound treatment that heats the cancer within the bone, relieves pain and improves function for most patients when other treatment options are limited. (MD Hurwitz et al., Mag-netic Resonance–Guided Focused Ultrasound for Patients With Pain-ful Bone Metastases: Phase III Trial Results, J Natl Cancer Inst, 2014; doi: 10.1093). This is the first phase III study to use this technology in the treatment of cancer, “ says the

IMAGING NeWS


study’s principal investigator and lead author Dr M. Hurwitz of the department of Radiation Oncol-ogy at Thomas Jefferson Univer-sity. Although radiation therapy is commonly used to treat bone-related pain and is effective for most patients, not all patients experience pain relief and over time those who do may have recurrence of pain. In addition, it’s possible for a patient to receive the maximum radia-tion dose that can be safely deliv-ered without fully controlling the pain. In situations where radiation therapy is not an option, alternative treatments are required. A total of 147 patients from 17 centers in the U.S., Canada, Israel, Italy, and Rus-sia were enrolled in the study and randomized to undergo MRgFUS or a placebo treatment. Patients in the treatment group received focused ultrasound precisely tar-geted to their bone tumors to heat the tumor tissue to between 65 and 85 degrees Celsius, resulting in its destruction. During each treatment, the patients were monitored real-time via magnetic resonance imag-ing (MRI) to ensure the correct

tissue was targeted and the right temperatures were reached while ensuring heat in surrounding nor-mal tissues and organs remained at safe levels. Patients who did not respond to the placebo treatment within two weeks were allowed to be unblinded and offered MRgFUS. Patients responded well to treat-ment, with 64 percent experienc-ing either no pain or a significant reduction in their pain at three months as measured by a 2 point or greater decrease in the numeric rat-ing score (NRS) for pain, the clini-cally validated measurement tool. Many patients were able to reduce or stop use of opiod medications. Notably, most patients experienced pain relief and improved function-ing within several days of treatment.http://tinyurl.com/Hurwitz-et-al-paper

resting state fMri brain scans detect early parkinson’s Disease

Oxford University researchers have shown that resting state fMRI offers promise for early diagnosis of Parkinson’s disease. The team dem-onstrated that their fMRI can detect people who have early-stage Par-kinson’s disease with 85% accuracy, according to research published in Neurology, the medical journal of the American Academy of Neurol-ogy. (Szewczyk-Krolikowski K et al. Functional connectivity in the basal ganglia network differentiates PD patients from controls. Neurol-ogy. 2014 Jun 11. pii: 10.1212). Dr Michele Hu of the Nuffield Depart-ment of Clinical Neurosciences is one of the joint lead researchers and said: ‘We think that our RS fMRI test will be relevant for diagnosis of Par-kinson’s. We tested it in people with early-stage Parkinson’s. But because it is so sensitive in these patients, we hope it will be able to predict who is at risk of disease before any symptoms have developed. How-ever, this is something that we still have to show by further research. This new research takes us one step closer to diagnosing Parkinson’s at

a much earlier stage – one of the biggest challenges facing research into the condition”. Affecting 1 in 500 people, Parkinson’s disease is caused by the progressive loss of a particular set of nerve cells in the brain, but this damage to nerve cells will have been going on for a long time before symptoms become apparent.

Conventional MRI cannot detect early signs of Parkinson’s, so the Oxford researchers used resting-state fMRI, in which people are simply required to stay still in the scanner.

The team compared 19 people with early-stage Parkinson’s disease while not on medication with 19 healthy people, matched for age and gender. They found that the Parkin-son’s patients had much lower con-nectivity in the basal ganglia. The researchers were able to define a cut-off or threshold level of connectivity. Falling below this level was able to predict who had Parkinson’s disease with 100% sensitivity (it picked up everyone with Parkinson’s) and 89.5% specificity.

The scientists applied their MRI test to a second group of 13 early-stage Parkinson’s patients as a validation of the approach. They correctly identi-fied 11 out of the 13 patients (85% accuracy).

The researchers are now carrying out further studies of their MRI tech-nique with people who are at increased risk of Parkinson’s.http://tinyurl.com/Szewczyk-Krolikowski-K-paper

www.raysafe.com

RaySafe X2 with its new touch screen interface is groundbreaking in its performance as well as its simplicity. All parameters and waveforms are captured in one simple step. Just swipe the screen to enlarge and analyze. The new sensors are orientation independent and don’t require any selections or corrections.

Call us for a demonstration on how we can make your life so much easier...

Well, at least your X-ray measurements.

Contact us to learn more!+49 731 175 492

RaySafe X2LESS EFFORT. MORE INSIGHT.

www.raysafe.com

RaySafe X2 with its new touch screen interface is groundbreaking in its performance as well as its simplicity. All parameters and waveforms are captured in one simple step. Just swipe the screen to enlarge and analyze. The new sensors are orientation independent and don’t require any selections or corrections.

Call us for a demonstration on how we can make your life so much easier...

Well, at least your X-ray measurements.

Contact us to learn more!+49 731 175 492

RaySafe X2LESS EFFORT. MORE INSIGHT.

FIeLD MRIULtrA HIGH


ensuring a continuing supply of magnets for ultra high-field 7 Tesla MRI systemsAt the recent ISMRM/ESMRMB meeting in Milan, Siemens announced that in future they would be manufacturing their own magnets for use in ultrahigh-field 7 Tesla MRI systems. This follows the decision a year ago by Agilent to stop ultrahigh-field magnet production. Until now Agilent had built and supplied such magnets to Siemens on an OEM basis.

Diagnostic Imaging Europe talked to Bernd Ohnesorge, CEO of Siemens’ MR Business Unit, about the significance and implications of Siemens’ decision to manufacture their own 7 Tesla magnets for human whole-body MRI systems in-house.

Q First some background. Is it true that the decision taken nearly a year ago by Agilent to phase out production of

the magnets that they supplied on an OEM basis to MRI manu-facturers only affected magnets for ultrahigh-field MRI systems? Thus, MRI systems of low to medium field strength which are the work-horses in clinical applications are unaffected by the change in Agilent’s policy?That’s true. The magnets that Agilent manufactured and sup-plied to us are those that are used in MRI systems of field strength of 7 Tesla and above. The magnet supply for clinical MRI systems with 1.5 Tesla or 3 Tesla was not affected by Agi-lent’s decision, as these magnets tend to be manufactured by the major clinical MRI manufacturers themselves. And let me point out that the ultrahigh-field magnets we’re talking about here are solely for use in MRI systems designed for human imaging. There exists a range of ultrahigh-field MRI systems for use in purely experimental applications such as in small animal research. Siemens is not involved in this field.

Q Presumably Agilent decided to stop their production of high field magnets because they couldn’t justify continuing

production for financial or other economic reasons. What makes you think that Siemens can succeed where Agilent couldn’t?Of course I can’t comment on the precise details and reasons behind Agilent’s decision to stop the production of ultrahigh-field magnets for human use in the future. For that you’ll have to ask them. At Siemens, we were pleased with Agilent’s commitment to honor existing supply and service agreements which enables us to continue to deliver 7 Tesla MRI systems based on Agilent magnets for ongoing projects. Thereafter, we will phase in our own 7 Tesla magnets, which will be developed and produced at Siemens Magnet Technologies in Oxford, UK. In our premises at Oxford, we will be able to leverage our tech-nical know-how which has been built up over many years and also utilize our very modern infrastructure for the large-scale production of clinical 1.5 Tesla and 3 Tesla magnets.

Q So after all, it was an easy decision for you to opt for pro-ducing your own high-field magnets?

Well, I wouldn’t say that. Manufacturing high-field magnets is a technologically challenging field and it’s not one that you blunder into without careful prior consideration. It requires significant investments so, as you would expect, we carried out extensive feasibility studies and technical pre-developments on the way to make this decision. However, in the end it actually boiled down to a simple factor. Namely the commitment of Siemens as a major supplier of imaging equipment and of MRI scanners in general — and to us that includes 7 Tesla systems. We believe in the potential of 7 Tesla MRI for research and clinical application in humans and are committed to drive the development in the 7 Tesla field and to help stimulate its growth in the future. And this is not just wishful thinking, but based on the exciting and highly promising results already being generated by the very innovative group of people and the world leading institutions who currently use 7 Tesla MRI systems for human research purposes.

Q In practical terms how will your decision to get into high field magnet manufacturing play out? What about timing?

The home of our 7 Tesla magnet development and manufac-turing will be at Siemens Magnet Technology Ltd. (SMT) in Oxford, UK, which has been highly active in magnet develop-ment over the past years and has been responsible for the intro-duction of several significant innovations in MRI magnet tech-nology such as actively shielded magnets and re-condensing magnets with zero Helium boil-off. We feel that, given SMT’s leading position in the field of clinical magnets, there will be synergies between their existing activities and their future role in producing 7 Tesla magnets. At both, the level of the magnet and also at the level of complete systems, a large part of the components which make up a state-of-the-art clinical MRI sys-tem are modular so we expect further synergies between our existing production models and the in-house 7 Tesla systems.Having said that, we are very conscious about the technological

Bernd Ohnesorge is CEO of Siemens MRI division. When Agilent announced a year ago that they would gradually phase out produc-tion of high-field magnets for use in OEM production of 7T MRI scanners, Bernd pledged to evaluate alll pos-sibilities to enable continuation of the production of 7T MRI systems. With the recent announcement that Siemens would start their own in-house production of 7T magnets the promise has been fulfilled. This decision is being seen as a sign of Siemens’ continued commitment to 7T systems and MRI in general.


challenges involved with a 7 Tesla magnet development and are very confident to master them with the experience we have. Proceeding from SMT’s current production of 1.5 Tesla and 3 Tesla systems to magnets for 7 Tesla systems is a leap in terms of design and manufacturing. However part and parcel of the basic decision to go for our own 7 Tesla production is the understanding that we are in an excellent position to support this development with the appropriate technical and personnel resources to meet the challenge. With our engineering and project management back-ground we at Siemens are not afraid of using our long experience in the MRI field to push back technical frontiers. We have already set up R&D teams to optimize the design and manufacturing aspects for the new 7 Tesla magnet. Regarding timing we expect to be fully up and running with our 7 Tesla magnet production in about three years. This does not mean that in the intervening period everything will grind to a halt from the point of view of supply of 7 Tesla magnets. Siemens has a remaining pipeline of pre-ordered magnets from Agilent, based on which, we will continue to support ongoing 7 Tesla projects. In addition, our 7 Tesla system design is so modular that it can also be integrated onto other ultrahigh-field OEM magnets, and this for already installed ones and also for new ones if there would be any other source become available.

Q You mentioned that there was an exclusive club of people already using 7 Tesla MRI in human research applications.

How many 7T installations are there?Right now there are more than 40 Siemens ultrahigh-field systems for human use installed or on order throughout the world, 7 Tesla

systems and systems with more than 7 Tesla combined. The current footprint is mostly in Europe, and in North America but there are more and more installations in place and in progress in particular in Asia Pacific, including China, Korea, Japan and Australia, and also in South America. We are proud that nearly two third of the current customer base of ultrahigh-field MR systems are partnering with Siemens and we strive to sustain or even further develop our position in this field, as our own 7 Tesla magnet is coming to the market. One other practical effect of this leadership is for example the relatively short installation time of new 7 Tesla systems. Despite their technological complexity they are typically up and running within two to three months, generating high-quality and reproducible results by using well-established protocols and quali-fied local coils. As I said, our existing strong presence in the 7 Tesla MRI field, was part of our decision to go our own way with in-house 7 Tesla magnet production. We feel not just a commitment to the field as it stands today but we also remain convinced of its future potential. Actually when calling the group of current 7 Tesla users an “exlusive club”, this should not imply that the exclusiveness means unwillingness to share results. On the contrary, not only is there a growing number of 7 Tesla publications each year but, in addition, we at Siemens organize regular Ultra High Field (UHF) users meetings, whose objective is to facilitate the exchange of all experiences, knowledge and research results. This is based on the recognition that it is in everybody’s interest to share what has been gained from working in this exciting but non-routine and experi-mental environment. A practical example of this is the exchange of already developed, fully validated protocols to other users who can then employ them “straight out of the box”.

Q On which biological/clinica areas is this ultrahigh-field MR research focused?

As could be expected, neurology applications is a particularly active field with fMRI studies of basic brain function and patholo-gies such as neurodegenerative diseases being particularly inten-sively studied. However it’s not limited to that; there is on-going activity and a growing number of publications in the musculoskel-etal field in the examination of joints such as knees. Research is also being carried out in cardiac and other vascular applications. The high resolution of ultrahigh-field MR images is of particular relevance in all these applications. However I should point out that this is still all research work: the ultrahigh-field MR systems are at this point not approved by the regulatory authorities for patient examinations and diagnostic use. From the regulatory point of view, approval of 7 Tesla MR systems for clinical diag-nostic purposes may be possible in some years from now as safety of the systems at high field strength is being proven. At the same time, some questions are still to be answered concerning the clinical added value but also regarding cost-benefit analysis and reimbursement possibilities. However, the only way to be able to address these questions is through the steady supply of research results and data from clinical studies using 7 Tesla systems. We see our continuing role in the 7 Tesla MRI area as providing the hardware and software of 7 Tesla systems as well as our know-how to researchers who generate such data, to reach the point where submissions for regulatory approval of 7 Tesla MRI for use in clini-cal patient examinations may become possible.

An elite club of Ultra High Field MRI users throughout the world investigate research applications in areas such as neurology but increasingly in musculoskeletal applica-tions such as joint examinations, where the extremely high resolution provided by the new systems is particularly useful. Nearly 70% of all such UHF researchers use Siemens systems.

in the future, magnets for siemens 7T Mri systems will be developed and manufactured in siemens Magnet Technology (sMT) facility located in oxford, UK, where currently lower field strength systems are produced. it is expected that the use of the existing infrastructure at sMT will allow significant manu-facturing synergies.

eDuCATIONrAdIoLoGY


By Drs A Vollman, R Hulen, S Dulchavsky & M van Holsbeeck

Fusing MRI with ultrasound images — a tool to enhance ultrasound education and improve clinical ultrasound

The clinical use of Musculoskeletal (MSK) ultra-sound (US) has increased significantly largely due to its diagnostic accuracy for many indications and its significantly lower cost than other cross-sectional imaging modalities. In the United States in 2009 radi-ologists interpreted 91,022 MSK ultrasound studies compared with only 40,877 in 2000 [1]. MSK ultra-sound has excellent resolution for soft tissue applica-tions, making it a good choice to diagnose conditions such as partial or full thickness tendon tears, nerve entrapments, muscle strains, ligament strains, joint effusions, etc [2]. However, despite its rising use, there is a relative lack of formal training in MSK ultrasound for both residents and fellows, leaving few musculo-skeletal radiologists with specific expertise compared to their generally more robust experience in musculo-skeletal MRI.

One of the biggest challenges for radiologists with-out formal training in MSK ultrasound is the lack of anatomic orientation when interpreting MSK ultra-sound. The field of view can be narrow with few recognizable anatomical landmarks, a limitation that can undermine confidence with MSK ultrasound diag-nosis. New commercially available fusion ultrasound technology (vendors include but are not limited to Hitachi, Toshiba, Siemens, Philips, General Electric) offers a potential solution for these issues and may also provide excellent teaching opportunities for MSK ultrasound and ultrasound at large. The technology uses MRI (or CT) to provide a larger field of view and easier identifiable relationships while maintaining the benefits of the cost effective, low radiation, and real time scan of the ultrasound device.

FUsING ULtrAsoUNd ANd MrI—How It worksIn order to fuse MRI and US images, the patient must have had a prior MRI which can be uploaded onto the fusion ultrasound machine. This MRI can be synchronized with a real time US image. This linkage

is accomplished by selecting 3 or more easily found landmarks on both the MRI and US images and lock-ing them in the appropriate plane. Once linked, elec-tromagnetic tracking sensors which can be attached to the ultrasound probe send positional data to an RF transmitter that is stationed in a fixed location next to the patient during the procedure. A GPS-like technol-ogy sends data back to the fusion US machine from the probe. On the US monitor, there will be a side-by-side projection of the real time ultrasound image and the correlating MRI. As the user moves the ultrasound probe, the MRI projection will automatically adjust showing the corresponding MR images in any plane the US probe is moved. Studies have shown that fusion of ultrasound to computed tomography (CT) is accurate, with a mean registration error of 0.3 mm, when performing ultrasound-guided injections into the sacroiliac joint of cadavers (3,4).

A stUdY deMoNstrAtING beNeFIts oF FUsIoN tecHNoLoGYA recent study carried out by our team, investigated whether the fusion of musculoskeletal MRI with their correlating ultrasound image would enhance both medical school students and/or radiology residents ability to identify more anatomy correctly then if just given an US image alone [5]. The study did prove that, with the addition of a MRI fused with an US image compared to the US image alone, radiology residents were able to answer more anatomy questions cor-rectly (p-value < 0.001). However, the medical school students did not benefit from having fused images. The conclusion of our study was that the radiology residents’ interpretation of the anatomy was enhanced by having the correlating MRI projected next to the US image. We theorized that the explanation behind

the author

Andrew Vollman, Md, rachel Hulen, Md, & Marnix van Holsbeeck, MD are at the Department of Radiology, henry Ford health System, Detroit, MI, USA

Scott Dulchavsky, MD, PhD, is at the Department of Surgery, henry Ford health System, Detroit, MI, USA

email: [email protected] FigUre 1. Ultrasound image of the anterior femoral head on an axial supine pelvis (left) fused with a proton density axial MRI image of the hip joint (right).


these findings was based on the prior knowledge of MRI anatomy. This allowed the residents to use the fused MRI as a beneficial tool when identi-fying anatomy correctly on the ultra-sound images.

FUsIoN ULtrAsoUNd’s pLAce IN rAdIoLoGY cUrrIcULUM Fusing technology offers the poten-tial to reshape the way ultrasound is taught in radiology residency pro-grams and fellowships (i.e. MSK) both during the acquiring image phase (scanning) and the interpretation of a prior acquired image. A suggested ultrasound teaching model could be using the fusion US device in sim-ulation labs. “Normal” MRIs can be uploaded on the fusion ultrasound machine. Using volunteer subjects, a resident or fellow would be able to perform fusion ultrasound scans and more easily verify their plane, land-marks and anatomy using the corre-

lating MRI (or CT). By performing these studies in a simulation center, they could practice finding landmarks and difficult anatomy over and over without being bothered by patient interference and overall could master the US anatomy. After the users have become more confident with the rec-ognition of the ultrasound anatomy and landmarks, they could then pro-ceed to use only the ultrasound device (without the assistance of the fusion technology) to perform their scans.

FUsIoN ULtrAsoUNd IN cLINIcAL prActIceThere are vast opportunities that are afforded by introducing fusion ultra-sound technology into the clinical setting. Galiano showed that having CT to verify landmarks on sonograms helped when performing facet joint and periarticular injections [6]. Avail-ability of fusion imaging may also entice more radiologists to perform

joint injections and tumor biopsies with ultrasound guidance rather than CT to guide the procedure [7)]. In addition, fusion imaging offers the possibility of monitoring chronic dis-ease with serial ultrasound fused to a baseline MRI or CT for anatomy, decreasing the high cost and radiation exposure associated with repeated CT surveillance [8].

coNcLUsIoN Radiologist or physicians with exten-sive training in cross sectional anat-omy (CT/MRI) may be the greatest beneficiaries of this new technology. Ultrasound fusion is an exciting new imaging technique that has the poten-tial to reshape attitudes toward MSK ultrasound and ultrasound in general, but more research is needed in order to assess its full utility. This technol-ogy shows promise as an important teaching tool that may help radiology residents and fellows use their emerg-ing knowledge of other cross sectional modalities to attain confidence and accuracy as they approach difficult ultrasound anatomy like MSK.

reFereNces1. Sharpe RE, nazarian ln, Parker l, et al.

Dramatically increased musculoskeletal ultra-sound utilization from 2000 to 2009, espe-cially by podiatrists in private offices. J Am coll Radiol. 2012 Feb;9:141.

2. Smith J, Finnoff JT. Diagnostic and interven-tional musculoskeletal ultrasound: part 1 fun-damentals. PM&R 2009; 1: 64.

3. Klauser AS, Peetrons P. Developments in mus-culoskeletal ultrasound and clinical applica-tions. Skeletal Radiol 2010; 39: 1061.

4. Klauser AS, De Zordo T, Feuchtner GM et al. Fusion of real-time US with cT images to guide sacroiliac joint injection in vitro and in vivo. Radiology. 2010 Aug; 256: 547.

5. Vollman A, hulen R, Dulchavsky S et al. Educational benefits of fusing magnetic resonance imaging with sonograms. J clin Ultrasound. 2014 Jun;42(5):257.

6. Galiano K. Obwgeseser AA, Bale R, et al. Ultrasound-guided and cT-navigation-assisted periarticular and facet joint injections in the lumbar and cervical spine: a new teaching tool to recognize the sonoanatomic pattern. Reg Anesth Pain Med. 2007; 32:254.

7. crocetti l, lencioni R, Debeni S et al. Targeting liver lesions for radiofrequency ablation: an experimental feasibility study using a cT-US fusion imaging system. Invest Radiol. 2008; 43 :33.

8. Brenner D, hall E. computed Tomography — An Increasing Source of Radiation Exposure. n Engl J Med 2007; 357:2277

FigUre 2. Ultrasound image of the posterior shoulder in prone positioning in the axial plane visualizing the spinoglenoid groove (left) with a correlating axial proton-density MRI image of the shoulder joint (right)

FigUre 3. Ultrasound image in the axial plane of the posterior hip region visualizing the gluteus maximus muscle (left) with a correlating MRI image of the posterior pelvis

TOMOGRAphYcoMpUted


personalizing an integrated approach to CT imagingComputed Tomography (CT) is the back-bone and work-horse of modern medi-cal imaging. Reflecting this central role, the numbers of CT exams carried out throughout the world continue to grow, and the technique itself continues to evolve. The rigid, deliberately standardized protocols of the past have given way to a recognition that personalization of the technique is advantageous for all concerned, with the potential to generate diagnostic images at ever lower radia-tion doses. Succesful personalization however requires an understanding and mastery of many interacting parameters, including scanner parameters such as tube voltage and current setting as well as contrast medium parameters such as concentration, injection flow rate, etc.

Exactly how best to optimize CT imaging was explained by three experts in the field to a spell-bound audience at a Bracco-sponsored symposium during the recent ECR meeting. This article describes the key messages of the symposium.

persoNALIzed ct IMAGING : Are we tHere Yet?By Dr. M FranconeThe concept of personalized med-icine was introduced more than a decade ago. The basis is a patient-centric approach in which medi-cal decisions, diagnostic tests and therapies are modulated to the genetic and clinical characteris-tics of the subject. As interest in personalized medicine contin-ues to increase, radiologists will play an increasingly important role in this process, particularly as regards the diagnostic pathway

and imaging-guided therapy. For computed tomography (CT) in particular, the

concept of a “standardized” or fixed protocol can no longer be applied given the issue of radiation expo-sure, the need for iodinated contrast agents and the fact that notable differences (and therefore “proce-dural” risks) exist between patients (e.g. children, elderly persons and subjects with chronic and/or neoplastic disease).

Final CT image quality depends on a complex combination of individual-, scanning- and contrast-related factors that are strictly inter-dependent and which need to be understood by CT users in order to fully optimize the acquisition and to tailor the examination to the specific clinical request.

coNtrAst cHArActerIstIcs ANd kINetIcs:The degree of contrast enhancement in CT is directly related to the amount of iodine accumulated within an organ/vessel, which physically leads to higher x-ray absorp-tion and therefore increases in CT attenuation.

Keeping in mind this very simple rule, it is important to recognize that vascular- and parenchymal enhancement is affected by different kinetics.

Vascular enhancement is mostly determined by the rela-tionship between the iodine administration per unit time (the iodine flux (mg I/s) versus the blood flow per unit time (i.e. the cardiac output [l/min]: ) whereas parenchy-mal enhancement is more dependent on the relationship between the total iodine load (mg I) and total volume of distribution (i.e. body weight [kg]).

A second important concept to remember when deal-ing with contrast optimization is that CT attenuation also strongly depends on the tube voltage applied.

Lowering x-ray tube voltage reduces the mean energy of photons interacting with the detector array leading to a rapid attenuation increase of elements with high atomic number such as calcium and/or iodine due to higher pho-toelectric interactions.

An average enhancement increase of ≈30% occurs when reducing the tube voltage from 120 kVp to 80 kVp while the effective radiation dose to the patient is reduced in parallel by approximately 70% on average.

pAtIeNts’ cHArActerIstIcs:Regardless of the physicochemical features and pharmaco-

M. Francone MD, PhD Dept of Radiological, Oncological & Path-ological SciencesSapienza University of Rome, Italy.


kinetics of contrast media, some key points related to patients’ individual characteristics have to be taken into consideration.

1. When administering iodinated contrast media there are several variables to be considered. Some are controllable, e.g. the iodine concentration, total amount and flow rate of the agent and some are not, e.g. patient body weight and cardiac output. For example, the contrast media transit time from the intrave-nous injection site to the beginning of the scan volume varies substantially across patients. It has been shown that, in 95% of patients, the contrast arrival time between the injection site and the abdominal aorta ranges between 14 and 28 s. To account for this variability, scan delays should be individualized. This can be accomplished either by using a test bolus or an automated bolus-triggering technique. Most vendors now have automated bolus triggering. In our experience, we have found this to be the most efficient and practical.

2. Cardiac output is inversely related to the degree of arterial enhancement, particularly in first pass dynamics. Because many patients referred for an MDCT angiography examination have coexisting cardiovascular disease with reduced cardiac output, the initial enhancement profiles may be shallow. Thus, for an MDCT angiography study of the peripheral vessels, if the CT acquisition is initiated too soon after the arrival of the con-trast medium bolus the opacification may not yet have reached adequate enhancement levels. An increase in the initial injection flow rate (e.g. 5–6 mL/s for the first 5 s of the injection) or the use of a high-concentration contrast medium (i.e. 400 mgI/mL) are options for minimizing this problem.

3. The patient’s weight may also have a significant impact on contrast enhancement. The relationship between maximum enhancement and body weight is linear, with a decrease in the degree of enhancement obtained as the body weight increases. For this reason, weight-optimized protocols have been pro-posed, mostly in countries where obesity is a major problem, in order to adjust the iodine dose to the individual weight of the patient. For example, if the total contrast media volumes are chosen relative to body weight, then 1.5–2.0 mL/kg body weight (450–600 mg I/kg body weight) is a reasonable quantity for MDCT angiography. However, it is important to emphasize

that body mass affects recirculation and vascular enhancement much more than first pass dynamics.

4. Other parameters affecting enhancement are a temporarily diminished venous return, which might occur following a forced Valsalva maneuver as a consequence of an exaggerated inspira-tion during a breath hold acquisition or by congenital heart dis-eases such as patent foramen ovale or an inter-ventricular defect that causes a temporary right-to-left shunt with early arterial enhancement. Even a patient’s height might potentially represent a relevant variable that influences contrast enhancement with an inverse and non-linear correlation with aortic attenuation reflect-ing the proportional increase of blood volume in taller subjects.

AcqUIsItIoN tecHNIqUe:Together with scanning parameters, acquisition parameters also play a crucial role in determining contrast administration strategies and thus have an obvious direct influence on final image quality.

Acquisition parameters affecting contrast enhancement include the following:

1. Scan duration which is influenced by the scanner type, pitch and acquisition mode selected (i.e. spiral vs. helical) and which directly influences the contrast administration strat-egy. The general assumption is that a long acquisition window requires a proportionally long injection time. Chasing the con-trast material with saline may not be feasible with slow CT scan-ners with the result that significant enhancement differences exist within the acquisition direction. Conversely, scanning too fast may overcome the contrast bolus leading to poor vascular/organ opacification.

2. Determination of scan direction which usually follows the flow direction (i.e. cranio-caudal) with the exceptions of pulmo-nary and carotid CT angiography. They require careful attention to slightly prolong the injection of contrast material to achieve adequate enhancement of upstream structures.

3. Definition of bolus scan arrival using either bolus test or bolus triggering techniques.

4. Definition of the optimal time-to-peak enhancement, which obviously depends on the contrast injection parameters (iodine concentration and flow) and the patient’s individual characteristics as discussed above.

5. Scan delay, representing the difference between enhance-ment peak and the start of the acquisition and which depends on the contrast injection duration and which should be optimized to obtain the highest enhancement within the target organ.

coNcLUsIoNs:Personalization of the examination protocol according to the patient’s individual characteristics is a necessary step in CT and contributes to better patient care. It also increases a radiologist’s involvement in the “personalized-medicine” process. This is currently a major target of diagnostic and therapeutic pro-cesses. Adequate tailoring of the acquisition technique requires knowledge of the impact on contrast enhancement of different individual-, scanner-, and contrast-related variables in order to synchronize image acquisition with the highest tissue enhance-ment within the scanned volume to maximize image quality.

Protocol optimization in a 56 year-old female undergoing coronary CT angiography (CCTA) due to recurrent atypical chest pain and inconclusive stress-ECG. CCTA was performed with a 1st generation dual source CT scanner with 70 mL of high iodine contrast agent (400 mgI/100 mL) + 40 mL of saline administered @ 5 mL/sec. The acquisition protocol was tailored to the patient’s personal characteristics and took into account her low body mass index (20.4) and heart rate (stable < 60 bpm). Radiation dose was minimized with a prospectively ECG-triggered acquisition technique applying a low tube current setting (90 Kvp) and processing raw data with iterative reconstructions for image noise reduction. Protocol optimization resulted in an extremely low effective dose of 0.8 mSv.


COMpuTeD TOMOGRAphYLow kV protocoLs – wHAt YoU sHoULd kNow

By Dr T Albrecht. Attenuation of iodine in con-trast media in CT depends on the mean photon energy used. The X-ray absorp-tion of iodine is at its maxi-mum when the mean energy of a p o l y c h r o -matic X-ray beam is close to the k-edge of iodine at

33.2 keV. The mean photon energy in the X-ray spectrum is 61.5 keV with 140 kVp, 56.8 keV with 120 kVp, 51.6 keV with 100 kVp, and 43.7 keV with 80 kVp respec-tively. Therefore, the use of lower tube voltage in CT (< 120 kVp) increases the mean attenuation value of iodine. Using 100 kVp or 80 kVp instead of 120 kVp in CT shifts the mean energy of the X-ray beam closer to the k-edge of iodine and thus improves contrast enhancement.

Reduction of tube voltage from 120 to 100 or 80 kVp, on the other hand, decreases radiation exposure to the patient. When reducing tube voltage, some compensatory increase in tube cur-rent is required to maintain image qual-ity. This is in the order of 50%, but the exact increase in tube current varies with body areas, different scanners and their automatic exposure control algorithms. The degree of radiation dose reduction, which occurs even after compensatory tube current increase, depends on a num-ber of factors including body habitus and body area under examination; it is more marked in thinner patients. When chang-ing from 120 to 100 kV, a dose reduc-tion of 20 – 50% is achieved in clinical practice; dose savings are most marked in the chest.

In summary, CT scanning at low tube voltage improves contrast enhancement and reduces patient radiation dose. In clinical practice, low kV CT scanning protocols can exploit these effects in sev-eral ways in comparison to standard 120 kVp protocols:

1. Increased contrast enhancement

with better contrast-to-noise-ratio at reduced radiation dose if the contrast volume remains unchanged

2. Constant contrast enhancement with slightly reduced contrast-to-noise-ratio (due to greater noise at low kVp) at reduced radiation dose if the contrast volume is considerably reduced

3. Ideally, a combined approach of the two options above is used: Slightly increased contrast enhancement with approximately equal contrast-to-noise-ratio at reduced radiation dose and a reduction of contrast volume by 15 – 20%.

AN INteGrAted ApproAcH: wHAt More Is Needed?By Prof. M Prokop

The over-all process of personaliza-tion involves several dis-tinct groups of activities. These are:

In div i du a l-ization.

Pro cess es that need to be individualized include the CT data acqui-

sition and reconstruction process itself, the setting of exposure parameters and the contrast media injection parameters. Integration. Interconnecting features that have to be integrated include voltage settings, (low kV) optimization of contrast media injection and iterative reconstruction. Management. All these individual aspects have to be managed efficiently. The management process includes recording and control of all protocols and ensuring that they are all up to date; recording their use and establishment of Key Performance Indicators (KPIs) Taking these items in turn, the message regarding data acquisition is “scan thin and read thick” i.e. collimation of 0.5 - 0.625 mm but reconstruction at 3 - 5mm. As for iterative reconstruction the message is clear “Use It”. People who don’t use iterative reconstruction but whose system enables it, should recon-sider their position. It is necessary how-

ever to have someone on the team who fully understands it and knows how to optimize it. As for exposure parameters most modern scanners make their opti-mization straightforward. So auto-matic exposure control should always be used, in conjunction with z-axis modulation which enables the dose to be modified for example as the scan passes from the shoulders to the neck. X-Y modulation is also possible on most modern scanners but it is NOT a good idea to use this without con-comitant z-axis modulation. Modula-tion is of course less effective with wide detectors. Regarding kVp adaptation, there are now automatic programs on the market. However even without these programs good judgment can adapt kVp accord-ing to how slim the patient is or in areas where there is very little absorption, etc.

For personalizing contrast injections there are of course many patient-dependent parameters involved but of these by far the most important are car-diac output and blood volume. Cardiac output is most affected by pregnancy (massive increase in cardiac output) and heart failure (massive decrease). Blood volume is generally correlated with patient size.Moving on to contrast injection param-eters the factors that determine how much enhancement is achieved are the iodine delivery rate (or flux) which basically affects arterial enhancement and the total iodine load which affects venous enhancement. Optimizing contrast enhancement is important for CNR. Contrast enhancement increases with smaller patients; lower cardiac output; lower kVp; higher iodine flux and higher iodine load. Typical increased enhancements achievable through reduction of kVp are of the order of 41% in changing from 120 kVp to 90 kVp. Alternatively for con-stant contrast enhancement lower kVp can mean a reduction of the amount of contrast media required, which can be of use in patients with poor renal function. One other way of looking at this is to use increased concentrations of contrast media to reduce the radia-tion dose. Enhancing the contrast in

Prof. Dr. med. Thomas Albrecht, FRCRInstitut für Radiologie und Interventionelle TherapieVivantes Klinikum Neukölln, BerlinGermany

Prof. Matthias Prokop is at Radboud University Nijmegen Medical Center Nijmegen, The Netherlands


the vessels means that we can accept a higher noise at the same signal to noise ratio. Thus if it were possible to increase the contrast used by 25% the dose could be reduced by about 35%. Thus, reduction in contrast at a con-stant CNR is more powerful in terms of reducing dose than a reduction in mA alone. Now we can bring iterative reconstruc-tion into all this. Iterative reconstruc-tion works best in structures where there is a high contrast gradient, so is very effective in the chest but much less effective in the brain. It also works well when there is a high signal to noise ratio in the arteries.However increases in contrast require adaptation of window settings. A standard window setting can give too much contrast and disturbing noise, whereas increasing the window width as a function of the increase in con-trast used gives a normal contrast and a reduction in the visible noise.

In practice however implementation of all this means that protocols have to be designed to personalize contrast injection. This means:• Introducing a contrast calculator based on patient weight (possibly including other factors such as height and sex). •Establishing protocols for low radiation

exposure, for example in CTA, cardiac, slim patients and children. This involves use of high concentration of contrast and high flow with a low kVp and a wide window together with iterative reconstruction.• Establishing protocols for low con-trast exposure. These would be useful in elderly patients or those with poor renal function. Such protocols involve low contrast flow and volume, a low kVp, a normal window and as always iterative reconstruction.

prActIcAL IMpLeMeNtAtIoN Contrast Administration.

In practice, however, personalization of contrast media dosing as described above — in other words giving different amounts of contrast to different patients presents a few problems.

One of these is simply the regula-tory requirements in various European countries, which can mean that in some countries it is almost impossible to give individual doses. A single syringe per patient is almost obligatory.

Another practical issue is the simple fact that there are fixed sizes of bottles, with the result that large amounts of contrast media may have to be unneces-sarily discarded.

Home-made systems to pro vide multi-patient systems for administering various doses of contrast media inevi-tably run into issues of maintenance of sterility and dealing with problems such as the back-flow of blood.

To meet all these challenges, Bracco has introduced a multi-dosing sys-tem, the CT Exprès contrast injector device, which does not use syringes but instead uses roller pumps. Fluid feeds come from saline infusion bags and two bottles of 500mL contrast media, and a “click” system for aseptically fitting the lines.

Practical experience of the system at Nijmegen confirmed the expected performance and showed the high degree of flexibility of the system. Up to 24 phases can be used and there is enough contrast for 8 - 12 patients. The contrast media is warmed, which is important when administering rela-tively large volumes of contrast media. Particularly appreciated by radiogra-phy technicians is the simplicity of use and the fact that the system is “clean”, with no spillage of contrast.

Protocol management This important aspect can become dif-ficult, particularly when there are sev-eral different types of scanner in the department. Protocol management is also important to monitor compliance of radiographers, for example when well-meant indi-vidual changes are introduced which may not always be to the benefit of the patient. Error detection systems must be implemented.Likewise just about every radiology department has some old, out-dated non-optimized protocols lurking around somewhere in the system. If these hap-pen to be resurrected and used there can be a significant risk of totally inap-propriate doses being administered and wrong amounts of contrast media being used. Finally, protocol management systems must include facilities for recording total consumption of contrast media and radiation exposures. To address all these points, Bracco has developed a contrast and dose manage-ment system, known as Nexo. The principal feature of Nexo is the cen-tralized protocol management facility and its control of the use of protocols. Data can be uploaded to and from the contrast injector system and also to the RIS/PACS/HIS systems. This fea-ture is really useful in the context of centralizing all relevant patient data in one place. Without an adequate proto-col management system inputting con-trast media data can be laborious and time-consuming.

Thus to summarize and to refer to the earlier question “ What more is needed?”, the answer is not much, especially since the introduction of the new tools. The key is to now implement them into routine clinical practice.

Compatible with a large array of contrast media bottles and volumes, the CT Exprès system enables an opti-mized approach to contrast media injection, with direct injection from contrast media bottles.

In addition to the centralized protocol management facility provided by Nexo, the system also enables the calculation of Key Performance Indicators.

IMAGINGbreAst


By Dr M Lobbes

Contrast-enhanced spectral mammography: a promising breast imaging tool in the evaluation of patients referred from breast cancer screening programsIn this article, we present a summary of our recent study, in which we evalu-ated the performance of Contrast-Enhanced Spectral Mammography (CESM) in women referred from a breast cancer screening program. We showed that CESM is a powerful problem-solving tool that enables false-positive findings to be ruled out quickly, and true-positive cases to be detected with a high degree of confidence.

In the Netherlands, a program for breast cancer screen-ing was introduced in 1989 and fully implemented in 1997 [1]. Women aged 50 to 75 years are invited to participate every two years. The screening program in the Nether-lands is organized and managed centrally by the National Expert and Training Centre for Breast Cancer Screen-ing, which also provides dedicated training and refresher courses for (future) screening radiologists. As a result, the Dutch screening program has one of the lowest reported referral rates, namely of approximately 2%. Even with this low referral rate, the prevalence of breast cancer of the referred women remains modest. In our own institute, dis-ease prevalence of this patient group is approximately 30%. Thus, 7 out of 10 referred women do not have breast cancer. In countries with an even higher referral rate the number of false-positive referrals is even higher.

Medical professionals should therefore realize that these screened women are in general healthy individuals. A notifi-cation that an abnormality had been found in the screening mammogram turns these women into patients. Although studies have claimed in the past that the experience of a false-positive screening exam affects women’s health, in their recent study Tosteston et al. did not report such findings [2]. However, they did observe short term anxiety about the test results, which rapidly declined over time.

However, as enshrined in our Hippocratic oath, doctors should do no harm in the first place. False-positive refer-rals are inherent to every screening program, but it is our responsibility as doctors to reassure these patients as soon as possible, with the highest degree of confidence and the least amount of additional exams or hospital visits.

Digital breast tomosynthesis (DBT) is currently being studied as an advanced breast imaging modality that might aid in the reduction of the number of false-positive recalls. Indeed Skaane et al. showed in a screening population that the number of such recalls is reduced by 15% when using DBT (versus mammography alone) [3]. However, the introduction of DBT into a screening program can suffer from some practical difficulties. The amount of data acquired from a DBT is much higher than a normal mam-mogram and needs to be stored on digital archives for many women over a long period of time. In addition, the radiologist’s reading time is doubled when using DBT [3]. Alternatively, DBT could be used as a problem-solving tool in inconclusive cases within the setting of the screening program, but its main limitation is that it can only rule out superposition densities caused by normal fibroglandular tissue. Consequently, DBT can be an alternative for the mammographic work-up. Many false-positive recalls are caused by cysts, in which DBT cannot rule out an underly-ing breast cancer. Additional conventional imaging, for example ultrasound, would still be needed.

We recently studied the performance of a new advanced mammography technique, contrast-enhanced spectral mammography (CESM), in women referred from the breast cancer screening program. We found that CESM, in contrast to DBT, could be a promising tool for use as an alternative for the entire conventional work-up [4].

the author

Marc bI Lobbes, Md, phd., is at

Department of Radiology and nuclear Medicine,

Maastricht University Medical center, P.O. Box 5800,

6202 AZ Maastricht,

The netherlands.

Tel : +31 43 387 4910.



coNtrAst eNHANced spectrAL MAMMoGrApHY In a CESM exam, an iodine-based contrast agent is intrave-

nously administered two minutes prior to the acquisition of the mammographic images. The contrast agents used are the same commercially available products which are also used in CT exams, for example. Thus, the same contra-indications apply for CESM as with any other radiologic exam using these agents. After contrast administration, CESM acquires a set of low- and high-energy images for both breasts in the two standard mam-mography views: craniocaudal and mediolateral oblique. Other, special views can also be carried out in the CESM technique.

The low-energy images are obtained using molybdenym (Mo) and rhodium (Rh) target and Mo and Rh filters at peak kilovoltage (kVp) values ranging from 26-31 kVp. As a result, the entire X-ray spectrum is above the k-edge of iodine, which is at 33.2 keV. The low-energy CESM image is comparable to a conventional mammogram.

Immediately after this first exposure, a high-energy image is acquired using a Mo target and a double layer filter (aluminium and copper) with peak kVp values ranging from 45-49 kVp (i.e. above the k-edge of iodine). This high-energy image contains

information on the presence of iodine, but is of no diagnostic value per se. It is used for the reconstruction of a so-called recombined image, in which enhancement of lesions is clearly visible [Figure 1]. In clinical practice, only the low-energy and recombined images are used for analysis [5].

The acquisition of these additional high-energy images is achieved in a few additional seconds, depending on compressed breast thickness. In between exposures, there is a delay of approximately 1 minute needed for the unit to process the images and the X-ray tube to cool down. The acquisition of the images takes approximately 7 minutes. The entire exam, includ-ing the placement and removal of the catheter needed for the contrast administration, takes approximately 20 minutes. Com-pared to mammography, the dose per exposure is increased by 81% with CESM, which is nevertheless still well within interna-tionally accepted ranges used in mammography [6].

cesM IN A popULAtIoN oF reLAtIVeLY Low dIseAse preVALeNce

Previously published studies showed that the diagnos-tic performance of CESM for breast cancer detection was superior to mammography alone [5], possibly even equal to breast MRI [7, 8]. However, these initial studies consisted of study populations enriched with breast cancer cases. As a result, the performance of CESM in populations with a limited disease prevalence still remained unclear. If CESM led to an unacceptable number of false-positive findings in these low-prevalence populations, the technique would still be regarded as not useful in clinical practice. For this reason, we studied the performance of CESM in a population with a lower breast cancer prevalence, namely women referred from the breast cancer screening program (i.e. a population with a disease prevalence of approximately 30%) [4].

In a period of 6 months, 116 women were judged eligible to undergo CESM for this indication. Three patients were excluded, leaving 113 patients to be included in the final analy-sis. A retrospective review of all cases was performed by two independent radiologists, who had first to score the low-energy exam (i.e. the conventional mammogram) using the BI-RADS classification score. Next, the entire CESM exam, including the recombined images, was presented. The radiologists were subsequently allowed to modify their previous score if they deemed it necessary. BI-RADS scores 1-3 were considered benign, whereas 4-5 were considered malignant. Using this cut-off, diagnostic performance parameters and area-under-the-ROC-curve (AUC) values could be calculated.

For the low-energy images, sensitivity, specificity, positive predictive value (PPV) and negative predictive value (NPV) were 97%, 42%, 40% and 97%, respectively. When CESM was added to the analysis, all parameters significantly increased to 100%, 88%, 76% and 100%, respectively. In short, breast cancer detection remained stable, but very interestingly, CESM was able to reliably rule out many false-positive findings. This is shown

FigUre 1. Typical example of a CESM exam. First, a low-energy image is acquired resembling a normal mammogram (left panel), immediately followed by the acqui-sition of a high-energy image, which is of no diagnostic value (middle panel). The low- and high-energy images are used to reconstruct the recombined image (right panel), in which enhancement of lesions can be seen. In this case, a multifocal invasive breast cancer was detected by CESM (white arrows).

FigUre 2. Summary of study results. When using cESM, sensitivity and nPV slightly increase. More interestingly, specificity and PPV are almost doubled. Image adapted from original [4] and reproduced courtesy of the publishers of European Radiology.

Cardiovascular and Interventional Radiological Society of EuropeC RSE

The CIRSE Annual Meeting is by far the largest and most signi� cant platform for minimally invasive, image-guided medicine, welcoming over 6,500 delegates from 94 countries in 2013.

CIRSE 2014 will o� er over 250 hours of scienti� c and educational content transmitted through a range of formats, including special focus sessions, hands-onworkshops, interactive case discussions, electronic posters, debates, a state-of-the-art technical exhibition and much more.

MAIN TOPICS• Vascular Interventions • Transcatheter Embolisation• Non-Vascular Interventions• Interventional Oncology• Neurointerventions• IR Management

For more information please visit us at:www.cirse.org

Check out highlights from past congresses and the latest Society updates here:www.esir.org – the largest educational platform in the fi eldfacebook.com/CIRSEsociety youtube.com/CIRSEsociety

September 13-17Glasgow, UK

CIRSE 2014

C RSEyears

Register Now!


BReAST IMAGING

by the improvement of specificity and PPV by a factor of more than 2 [Figures 2 and 3]. A similar trend was observed for the AUC values, which were 0.779 for the low-energy images and 0.976 for the entire CESM image sets (p<0.0001).

coNcLUsIoNWe have shown that CESM is superior

to normal mammography in breast can-cer detection even in populations with a lower breast cancer prevalence. More importantly, we showed that CESM can be an excellent problem-solving tool in women who are referred from breast can-cer screening programs.

Because of its potential in evalu-ating screening referrals, our depart-ment installed two dedicated CESM units which are used in everyday clinical practice for this indication. We have designed our workflow in such a way that all women coming

to our hospital for further analy-sis after a positive referral from the breast screening program immedi-ately have a CESM exam. Whenever it is necessary to check renal clearance (according to the ESUR Guidelines on Contrast Media), we do this using a rapid point-of-care method. Up until now, we have performed more than 360 CESM exams over 18 months and only observed three cases of reaction to the contrast agent, of which only one needed treatment.

This approach enables us to quickly rule out or diagnose breast cancer with a high degree of confidence and shows that CESM is feasible in everyday clinical practice. In our breast cancer institute, CESM has quickly established itself as a powerful new breast imaging tool.

reFereNces1. Timmers JMh, Van Doorne-nagtegaal hJ, Zonderland hM, et al. The Breast Imaging Reporting and Data System (BI-RADS) in the Dutch breast cancer screening

programme: its role as an assessment and stratifica-tion tool. Eur Radiol 2012; 22: 1717-1723.2. Tosteson An, Fryback DG, hammond cS, et al. consequences of false-positive screening mammo-grams. JAMA Intern Med 2014; Apr 21: Epub ahead of print.3. Skaane P, Bandos AI, Gullien R, et al. comparison of digital mammography alone and digital mammography plus tomosynthesis in a population-based screening program. Radiology 2013; 267: 47-56.4. lobbes MB, lalji U, houwers J,et al. contrast-enhanced spectral mammography in patients referred from the breast cancer screening programme. Eur Radiol 2014; Apr 3, Epub ahead of print.5. lobbes MB, Smidt Ml, houwers J, Tjan-heijnen Vc, Wildberger JE. contrast-enhanced mammogra-phy: techniques, current results, and potential indica-tions. clin Radiol 2013; 68: 935-944.6. Jeukens c, lalji U, Meijer E, et al. Radiation expo-sure of contrast-enhanced spectral mammography compared to full-field digital mammography. Invest Radiol 2014; in press.7. Jochelson MS, Dershaw DD, Sung JS, et al.Bilateral contrast-enhanced dual-energy digital mammography: feasibility and comparison with conventional digital mammography and MR imaging in women with known breast carcinoma. Radiology 2013; 266: 743-751.8. Fallenberg EM, Dromain c, Diekmann F, et al.. contrast-enhanced spectral mammography versus MRI: initial results in the detection of breast cancer and assessment of tumour size. Eur Radiol 2014; 24: 256-264.

FigUre 3. Example of the ability of CESM to confidently rule out false-positive referrals. This 51 year old patient was referred for a mass in the right breast, which clearly does not enhance on the recombined images and which shows a so-called ‘eclipse sign’, which is pathognomonic for a simple cyst (confirmed by targeted ultrasound and aspiration).

Cardiovascular and Interventional Radiological Society of EuropeC RSE

The CIRSE Annual Meeting is by far the largest and most signi� cant platform for minimally invasive, image-guided medicine, welcoming over 6,500 delegates from 94 countries in 2013.

CIRSE 2014 will o� er over 250 hours of scienti� c and educational content transmitted through a range of formats, including special focus sessions, hands-onworkshops, interactive case discussions, electronic posters, debates, a state-of-the-art technical exhibition and much more.

MAIN TOPICS• Vascular Interventions • Transcatheter Embolisation• Non-Vascular Interventions• Interventional Oncology• Neurointerventions• IR Management

For more information please visit us at:www.cirse.org

Check out highlights from past congresses and the latest Society updates here:www.esir.org – the largest educational platform in the fi eldfacebook.com/CIRSEsociety youtube.com/CIRSEsociety

September 13-17Glasgow, UK

CIRSE 2014

C RSEyears

Register Now!

A NEW GENERATION ULTRASOUND SYSTEM

Keen to experience this innovation in ultrasound?

www.tryarietta.com

www.hitachi-medical-systems.com


OuTLOOkINdUstrY

Game changing products that will shape radiology in the year aheadSome technologies are woven into the fabric of medical imaging, contributing diagnostically and logistically to countless millions of patient exams around the world. They have sustained medical imaging for years: ultrasound on the modality side; injectors on the periphery.

In the months ahead new products representing these technologies and built around them will shape the practice of medicine in Europe and the world, just as advances in breast imaging open new possibilities.

This article highlights the most striking of such new products presented at the ECR congress held in Vienna in March.

Visitors to the European Congress of Radiol-ogy in Vienna last March caught a glimpse of some of such new products and technologies

likely to shape the practice of medicine in the field of ultrasound.

Hitachi unveiled a new family of ultrasound scan-ners, namely the Arietta series, the first totally new products to come from the merger of Hitachi Medical and Aloka three years ago. Arietta 70 [Figure 1] the flagship of this new and soon-to-be growing family of scanners, supports the most advanced applications, including elastography and multimodality fusion. The Arietta 60 model is the workhorse, designed to support routine applications.

The new products reflect a design philosophy that blends the ergonomics of Aloka’s ProSound line with the advanced functionality and image quality of Hit-achi’s HI VISION ultrasound systems, according to company officials. Their Vienna unveiling was pre-ceded by a Japanese launch in January. But the ECR event signified the worldwide launch. CE mark and Japanese certification are in hand. FDA clearance is pending. First deliveries are expected in June. More additions to the family are sure to come.

Also in that European venue Guerbet introduced its FlowSens injector, featuring a unique soft bag injection system, for CT. [Figure 2] The new injector, designed by Medex, a Guerbet Group company, supports single and multiple patients. Simplicity and flexibility are

key design elements, according to the company, mak-ing FlowSens easy to use and exceptionally safe. It is compatible with all contrast media and saline products. Commercialization will begin this summer.

Other products with a likely impact on the European market and ultimately the practice of medical imaging were presented and showcased at the European Con-gress of Radiology.

The French innovator of Shearwave elastogra-phy, Supersonic Imagine, plans to leverage the use of this capability in concert with B Mode to help practitioners evaluate the liver [Figure 3] Shearwave elastography quantifies the elasticity of liver tissue,

the Author:

Greg Freiherr is head of the US-based firm, the Freiherr group, which specializes in corporate consulting and editorial services for the medical imaging industry email: [email protected]

FigUre 1. Arietta 70 is the new flagship of Hitachi Medical’s ultrasound line. First unveiled to an international audience at ECR 2014, the new product and its mid-tier sibling, Arietta 60, are the first of a new family of scanners to come out of the combined R&D efforts of Hitachi and Aloka, who merged three years ago.Images courtesy of Hitachi Medical.

By Greg Freiherr

A NEW GENERATION ULTRASOUND SYSTEM

Keen to experience this innovation in ultrasound?

www.tryarietta.com

www.hitachi-medical-systems.com


INDuSTRY OuTLOOk

promising insights into particularly difficult-to-diagnose disease states, specifically hepatitis B and C, which have long incubation periods. With early diagnosis, new drugs nearing the marketplace market might help those infected before they require transplants, according to the com-pany. Supersonic Imagine will do so through its Aixplorer system, which began as a specialty unit for elastog-raphy but has since evolved into a general purpose, high-performance product.

Breast tomosynthesis pioneer Hologic plans to extend the use of

its Selenia Dimensions into clinical realms common in the 2D world but not in 3D, at least not yet. Routinely applied diagnostically, Hologic’s tomo now is being groomed for screening with the promulgation of results such as those obtained by researchers at Oslo University Hospital Ullevaal in Oslo, Norway. At ECR 2014, Dr Per Skaane described a four-year pro-spective trial of tomo covering 24,901 screening exams, published in the January 24, 2014, issue of Radiology with the conclusion that the combi-nation of 2D images reconstructed from tomo data using Hologic’s C-View algorithm and breast tomo

is “adequate for routine clinical use when interpreting screening mam-mograms.” [Figure 4]. Continuing the evolution of tomo will be the addi-tion of biopsy, a capability already within Hologic’s reach. The compa-ny’s Affirm breast biopsy guidance system is designed to perform either stereotactic or tomosynthesis breast biopsy. Affirm can be mounted on any Selenia Dimensions gantry and its controls are integrated into the Selenia Dimensions acquisition workstation.

With both CE mark and FDA approval in hand, the company is also readying contrast mammography as part of tomosynthesis, a capability that promises to offer physiological infor-mation about the lesion to comple-ment the morphology provided by tomo.

For patients with dense breasts, MR may offer insights not otherwise possible. This modality is unabash-edly more sensitive to the signs of cancer than any other modality. The problem is its non- specificity. At ECR 2014, contrast media manufac-turer Bracco built the case for boost-ing MRI with its MultiHance prod-uct, which last year was approved for breast MRI by the Reference Member State (the UK Medicines and Health Care products Regula-tory Agency.

MultiHance is indicated “for the detection of malignant lesions in patients where breast cancer is known or suspected on the basis of previous mammography or ultra-sound results.” The company is pur-suing approvals in other European Concerned Member States using the regulatory approval process called “Mutual Recognition Procedure.” At the European Congress, Dr. Fiona Gilbert, head of Department Radiol-ogy at the University of Cambridge School of Clinical Medicine in Cam-bridge, UK, asserted that the sensitiv-ity of MR makes this modality the gold standard for cancer detection and that MR with Bracco’s Multi-Hance produced superior results in the detection of cancer compared to the conventional techniques, as well MR scans performed with Multi-Hance’s rival, Magnevist.

FigUre 2. Guerbet introduced its new FlowSens injector at the ECR 2014. Unlike competing models, the injector uses soft bags rather than vials of CT contrast media.Photo by Greg Freiherr

FigUre 3 At ECR 2014, Supersonic Imagine described how liver imaging with Shearwave elastography and B Mode on its Aixplorer may pick up early signs of liver disease, such as Hepatitis B and C. These disease states each affect the elasticity of the liver, which can be documented on Aixplorer.Photo by Greg Freiherr


“Ease of use” has become a cli-ché in medical imaging’s lingua franca. But this may be changing, as companies look seriously at new ways to do old jobs. Siemens’ Ysio Max DR system features redesigned detectors and novel software to do exams more efficiently. Shown first at RSNA 2013, Ysio Max DR debuted in Europe at ECR 2014 with the MAX wi-D, a light, thin, fast wireless 35 x 43 cm detector with a handle and MAX mini, a 24x30 cm wireless detector dedicated to orthopedic, pediatric, and trauma cases. The detectors can be inter-changed quickly using MAXswap, which offers one-click registration. MAXalign boosts productivity by displaying the detector angle on the system’s touch display, assuring sin-gle-shot chest images in bed.

Concern about patient dose will have far-ranging effects in the months and years ahead, ones that are only beginning to be felt. Dose manage-ment techniques built into CT proto-cols are being developed, with results that not only lower patient exposure to radiation but improve image quality. At ECR 2014 Prof. Thomas Albrecht from the Institut für Radiologie und Interventionelle Therapie Vivantes, Klinikum Neukölln in Berlin, Ger-many showed how CT, enhanced with contrast media and performed at 100 kVp or 80kVp instead of 120 kVp, can reduce patient radiation dose and improve contrast enhancement. Just

changing from 120kVp to 100 kVp can increase image contrast, while reduc-ing patient radiation dose by 20% to 50%, according to Albrecht, while improving image quality. The reason, he said, is that the iodine in contrast media absorbs more X rays, boosting contrast.

In CT, dose management has syn-ergized with several families of sys-tems that are ready to diffuse into the marketplace. At RSNA 2013, new flagship CT scanners appeared –Sie-mens’ Somatom Force, GE’s Revo-lution CT, and Philips IQon. Each exemplifies its company’s proprie-tary twist to high-end CT. Somatom Force leverages a dual-energy approach to improved visualization and characterization; IQon uses spectral CT, delivering colorized images indicating the composition of structures; the over-sized detector on GE’s Revolution enables whole-organ coverage. All offer low dose. Two promise to spread low dose, high performance to more than just the stratospheric tiers of CT.

At ECR 2014 Somatom Force and Revolution CT were framed as the parents of scanner families. Siemens introduced its 16-slice Somatom Scope for handling mainstream radiology procedures. GE unveiled the Revolu-tion GSI and Evo as mid-tier scanners, differentiating them from the high-end Revolution in terms of detector size rather than slices – 16 cm for the Revolution; 4 cm each for the newest family members.

Like its predecessor, Aquilion One (the scanner that in 2008 took CT to new heights of whole-organ acquisi-tions), Toshiba’s new Aquilion One Next Generation provides whole organ coverage in a single rotation. This Next Generation scanner, however, has a larger gantry aperture and new detec-tor, the Quantum VI, which the com-pany says offers a higher light output for optimized dose reduction.

These CTs are in the front line of medical practice, standard bearers for sophisticated technologies. But the technologies and protocols that enable them, as well as lower profile

products, are no less significant in the value they lend to medicine or how they will change its practice. They are part of a broader evolu-tion, one that will bring many dif-ferent twists and turns in the years ahead. Just as CT makers now are exploring divergent paths to visual-ize and characterize tissue, so too might these paths converge in the future, as has happened with the dif-fusion of breast tomosynthesis and its ongoing evolution from diagnos-tic to screening modality.

The coming together of technolo-gies is being expressed in different ways and in multiple venues. Again Vienna provided a showcase for new ideas. At ECR 2014, in a Joint Ses-sion of the ESR and European Fed-eration of Societies for Ultrasound in Medicine and Biology, presenters described the clinical advantages of combining ultrasound with MR and CT, as well as enhancing sono-graphic images with contrast media. The combination of MR and ultra-sound may be useful in prostate biopsy, in which transverse images of the prostate, obtained with 2D transrectal ultrasound, are reformat-ted into a 3D image and fused with MR images. MR-marked lesions are superimposed on the real-time ultra-sound image, allowing electromag-netic tracking of needle trajectories in targeted biopsies.

Similarly, ultrasound fusion with reformatted PET, CT and MR images might be used in the abdomen for the characterization and intervention of liver lesions, abscesses containing air and lesions poorly seen with ultra-sound. Ultrasound fusion may also be useful when evaluating lesions visual-ized on PET.

The integration of multiple modali-ties in diagnostic continuums, seen in women’s health with the selective uti-lization of ultrasound, mammography and MR, may come to characterize future medical practice.

The community has long recog-nized this complementarity. What has yet to be realized is how they can be expressed in a more coherent and effi-cient way.

FigUre 4. Synthetic 2D mammograms, created with Hologic’s C-View software, have been judged com-parable to digital mammograms. Such synthetic 2D images can be used with breast tomo when performing routine clinical screening mammography, according to research presented at ECR 2014. Images courtesy of Hologic

28 D I E U R O P E APRIL/MAY 2014

By Dr Dž. Belkić & Dr K. Belkić

The Role of Optimized Molecular Imaging in personalized Cancer MedicineMolecular imaging using PET/cT or magnetic resonance spectroscopy and chemical shift imaging provides added value compared to modalities of anatomic imaging alone, and yields higher specificity in identifying cancers. The absence of ionizing radiation means that molecu-lar imaging by magnetic resonance is particularly appropriate for the surveillance of persons at increased risk of certain solid cancers and for post-therapeutic monitoring of patients, espe-cially pediatric cases. Molecular imaging is important for “targeted therapies” in the detec-tion, monitoring and prediction of the widely varying responses of individual cancer patients, for example to the same type of drugs. In addition molecular imaging can guide the treatment of individual patients, thus enabling the practical implementation of the concept of “personalized cancer medicine”.

If recent advances in the analysis of molecular imaging data are judiciously incorporated into clinical practice, advances can be expected in all aspects of oncology, including earlier cancer detection especially for persons at high risk. This article describes how MR-based molecular imaging in which signal processing methods are mathematically optimized has the potential to improve early cancer detection.

MoLecULAr IMAGINGMolecular imaging integrates cellular and molecular biol-

ogy with diagnostic imaging. With the rapidly growing extent and sophistication in our understanding of the cell biology of cancer, molecular imaging offers a strategic, non-invasive bridge to clinical oncology. Not only can the presence of the cancer disease process be identified but its extent and sever-ity can also be assessed and followed over time. In addition, molecular imaging is invaluable for target definition in dose planning procedures in radiotherapy (RT) and for image-guided biopsy and surgery. Molecular targeting can also help determine the extent of certain tumors such as androgen-sen-sitive prostate cancers and neuroendocrine tumors. Overall, molecular imaging is becoming recognized as a key compo-nent of personalized cancer medicine [1, 2].

One of the advantages of molecular imaging is its potential to improve diagnostic accuracy for cancer. Anatomic imaging using magnetic resonance imaging (MRI), ultrasound and computerized tomography (CT) can be quite sensitive but are often non-specific, frequently resulting in “over-diagnosis” [3], with potential untoward consequences including unneces-sary workup, expense and psychological distress. Adherence to cancer screening guidelines and other medical indications may thereby be compromised.

posItroN eMIssIoN toMoGrApHY pLUs coMpUterIzed toMoGrApHY

Currently the molecular imaging modality that is most often used in clinical oncology is positron emission tomography (PET). When PET is combined with computerized tomography (PET-CT), both anatomic and molecular imaging information can be obtained, which has the potential of having a major impact on the care of patients with cancer. The main biomarker currently employed in PET is the glucose analog 18F-fluoro-2-deoxy-D-glucose (FDG), whose rationale for use is based on the fact that anaerobic glycolysis is enhanced in cancer cells. Compared to anatomical imaging alone, whole-body PET-CT improves cancer staging, and has been reported to have led in up to 40% of cases to changes in the therapeutic management of cancer patients [1, 4]. In this way, radiation treatment can be improved. In-beam PET-CT cameras can potentially monitor dose depositions in situ such that corrections could be made in subsequent dose deliveries, so long as the associated mathemati-cal data processing problems are solved. PET-CT cameras are currently being used in light ion radiation therapy in which the

the Authors:

professor dževad belkić, phd 1, Adjunct professor karen belkić, Md, phd 1-3,

are at

1. Department of Oncology/Pathology, Karolinska Institute, P.O. Box 260 Stockholm, SE-171776, Sweden

2. claremont Graduate University, School of community and Global health, claremont, california, USA

3. Institute for Prevention Research, the University of Southern california School of Medicine, los Angeles, california

correspondence to:

Professor Dževad Belkić, Department of Oncology-Pathology,

Karolinska Institute, Building P-9, Box 260, Stockholm SE-17176 Sweden


IMAGINGMoLecULAr

APRIL/MAY 2014 D I E U R O P E 29

primary particle beam generates secondary nuclei capable of positron emission (e.g. 12C is transmuted to the β+-emitters 11B and 10B). This type of in situ monitoring of dose dis-tributions could yield invaluable information about the patient’s response to irradiation. The response could be used to adjust the simulation codes and optimize dose delivery.

FDG-PET has also been used to identify areas of hyper-metabolism needing a boost dose [1]. The relative lack of specificity of FDG and other radiotracers continues to be a problem although the emergence of new molecular tracers may alleviate this issue.

tHe coMbINAtIoN oF MrI ANd Mr spectroscopY

Combined molecular and anatomic imaging can also be achieved by combin-ing magnetic resonance spectroscopy (MRS) with MRI to give magnetic resonance spec-troscopic imaging (MRSI). Since there is no exposure to ionizing radiation, repeated monitoring does not have the risk of induc-ing secondary cancers, which is especially important in pediatric cases and for persons at increased risk of cancer, in whom surveil-lance is indicated at more frequent intervals and starting at a younger age. MRS and MRSI can potentially provide information on sev-eral metabolites at the same time. Of these, choline, which resonates at ~3.2 ppm has been particularly useful as a reflection of the phospholipid metabolism in cell membranes, and as a marker of membrane damage, cel-lular proliferation and changes in cell density, all of which typically occur in malignancy.

Applications of MRSI have been most widespread in neuro-oncology, as well as for the initial detection and characterization of prostate cancer, staging, treatment planning, surveillance for residual cancer and local recurrence after therapy. Diagnosis of other cancers such as breast cancer, sarcoma, head and neck cancer has been aided by MRS/MRSI.

Magnetic resonance spectroscopic imag-ing has also been used quite extensively in radiation neuro-oncology and in radio-therapy of prostate cancer, e.g. for target definition, to identify high-grade areas for boost doses and for post-therapeutic follow-up. MRSI helps distinguish recurrent brain tumors and radiation necrosis post-RT. In this context, advanced signal processing could provide more effective assessment of the effects of RT, and evaluation of any new lesions that arise post-RT. Early detection of

post-therapeutic residual cancer by MRSI also has the potential of enabling more rapid intervention with a consequent improve-ment in patient outcome. The use of MRS and MRSI can also enable the response to chemotherapy to be assessed, so that prompt adjustments to the regimen can be made.

Most examinations in MRS/MRSI are con-cerned with the quantification of steady-state concentrations of metabolites of diagnostic significance. This can be complemented by dynamic MRS/MRSI studies using various prepolarized biomarkers to investigate the uptake kinetics and relaxation of time-depen-dent metabolite concentrations to their equi-librium stationary values. Such dynamic MRS/MRSI examinations have the potential to opti-mize both individual treatment strategies and to monitor treatment responses, as implied in the concept of “personalized medicine”.

Overall, the combination of MRS and MRSI complements MRI alone by yield-ing invaluable diagnostic information on the metabolism of the tumor tissue, often before changes can be identified reliably on conventional anatomical images. Detecting with confidence a small tumor by MRSI is of utmost importance, since it opens the possibility of selecting an appropriate treat-ment at an early stage in the disease.

Higher magnetic field systems are being used in some oncology centers with the aim of enhancing resolution and signal-to-noise ratio in MRS and MRSI. However, overlap-ping resonances deteriorate at higher field strengths compared to 1.5T. In addition, the cost of high field scanners is currently pro-hibitive for routine clinical use, especially for regular monitoring purposes.

AN ALterNAtIVe ApproAcHOur approach has been different, namely

the improvement of resolution in MRS and MRSI in 1.5 — 3T scanners via the use of powerful mathematical algorithms, specifi-cally the fast Padé transform (FPT) for opti-mization of signal processing. The FPT can provide quantitative information on at least 25 brain metabolites in in vivo MRS data from 1.5T scanners. Several of these metabolites are of particular relevance in neuro-oncology. Studies suggest that similar added diagnostic value could also be forthcoming in breast, prostate and ovarian cancers [5, 6]. In vivo MRS and MRSI have made important contri-butions to clinical oncology by relying upon only a few metabolites, or, in breast cancer on

just a single metabolite, namely total choline. This latter has been comprehensively reviewed [7]. The FPT could expand this restricted metabolite window and thus refine molecular imaging, potentially allowing “personalized molecular fingerprint of individual tumors, as a basis for novel treatment algorithms” [1: p 183].

poteNtIAL breAktHroUGH IN breAst cANcer dIAGNostIcs

Recent results in breast cancer dem-onstrate the potential of the approach. In malignant transformation in the breast, a “glycerophosphocholine to phosphocholine switch” occurs with over-expression of the enzyme choline kinase responsible for phos-phocholine (PC) synthesis [8,9]. Within the cytosine diphosphate-choline pathway, cho-line, PC and glycerophosphocholine can be detected in the proton magnetic resonance spectrum, using the fast Padé transform [5]. However, this cannot be achieved in the conventional way by which MRS signals are processed, that is by fitting envelope spectra from the fast Fourier transform.

wHY Is optIMIzAtIoN oF sIGNAL pro-cessING so crItIcAL IN Mrs?

A question which often comes to clini-cians’ minds is: “what is so important about signal processing in diagnostics?”

This is addressed in Figure 1. Let’s look at the top, upper panel (1). This is an MRS time signal of total length N = 2048 sampled points, constituting the input data consis-tent with the encoding from breast cancer, as reported in Ref [10]. Within this time sig-nal lies the desired information, namely the molecular content of the tissue. However, in its present form, it is completely unin-terpretable. Transformation into the com-plementary frequency domain is needed to visualize the spectrum. This mapping, achieved through mathematical transforms, is possible because time and frequency are “conjugate variables”. Currently, in clinical practice, this mapping is carried out by the FFT which is built into every MR scanner. Because this feature is built-in, the clinician never sees the original time signal. Instead, the spectrum is displayed directly leading to the common misconception that it is the MRS spectrum which is actually recorded. Another common assumption is that the way to improve the spectrum is through stronger magnetic fields.

30 D I E U R O P E APRIL/MAY 2014

MOLeCuLAR IMAGING

0 500 1000 1500 2000−0.8

−0.4

0

0.4

0.8INPUT DATA: TIME SIGNAL

N = 2048 (FULL LENGTH)

NOISE: σ = 0.0289 RMS

BW = 6 kHz, τ = 1/BW

(1) TIME ( in UNITS of SAMPLING RATE τ)

TIM

E S

IGN

AL W

AVEF

OR

M (

au) Input Time Signal as Damped Harmonic Oscillations

3.23.223.243.263.283.3

0

0.5

1

1.5

2

2.5

3PADE: FPT

NP = 1700 , N = 2048


(3.281, 0.900)

M−INS

(3.273 ppm, 2.800 μM/g)TAU

(3.251, 0.725)β−GLC

(3.232, 0.225)GPC

(3.221, 2.250)PE

(3.220, 0.300)PC

(3.212, 0.100)

CHO

(5) CHEMICAL SHIFT (ppm)

CO

NC

ENTR

ATIO

NS

( μ

M/g

)

Concentrations: Input (+) and Reconstructed (o)

3.23.223.243.263.283.3

0

50

100

150

200

250FOURIER: FFT

BROADENED & SHORTENED PEAKS

N = 2048



ENVE

LOPE

SPE

CTR

UM

(au

)

Absorption Total Shape Spectrum (Envelope)

3.163.183.23.223.243.263.283.33.323.34

0

2

4

6

8

10

TRUE METABOLITES

PC*


FWH

M (1

000*

ppm

)

Input: Poles (+), Reconstructed: Poles (o), Zeros ( • )

NP = 1700, N = 2048PADE: FPT NOISE: σ = 0.00289 RMS

3.23.223.243.263.283.3

0

50

100

150

200

250Absorption Total Shape Spectrum (Envelope)

PADE: FPTNP = 1700 , N = 2048


M−INS

TAU

β−GLC

GPC

PC+PE

CHO


ENVE

LOPE

SPE

CTR

UM

(au

)

3.163.183.23.223.243.263.283.33.323.34

0

2

4

6

8

10

TRUE METABOLITES

PC*


FWH

M (1

000*

ppm

)

Input: Poles (+), Reconstructed: Poles (o), Zeros ( • )

NP = 1700, N = 2048PADE: FPT NOISE: σ = 0.0289 RMS

σpanel (7) = 10*σpanel (6)

3.23.223.243.263.283.3

0

50

100

150

200

250Absorption Component Shape Spectra

PADE: FPTNP = 1700 , N = 2048


M−INS

TAU

β−GLC

GPC

PE

PCCHO


CO

MPO

NEN

T S

PEC

TRA

(au)

3.163.183.23.223.243.263.283.33.323.34

0

50

100

150

200

250

M−INS

TAU

β−GLC

GPC

PE

PCCHO


PEAK

HEI

GH

TS (

au)

Peak Heights: Input (+) and Reconstructed (o)

PADE: FPT

NP = 1700, N = 2048


MRS for BREAST CANCER by FAST PADE TRANSFORM (FPT): panels (3)−(8), FAST FOURIER TRANSFORM (FFT): panel (2)

APRIL/MAY 2014 D I E U R O P E 31

HIGH-resoLUtIoN wItHIN tHe FptProceeding to the next two panels (2) and

(3) in Figure 1, it can be immediately seen that the quality of these two spectra is very different. Panel (2) shows the Fourier spectrum generated from the MRS time signal on panel (1) associ-ated with breast cancer. All 2048 signal points are used, but not much can be determined from the few rough peaks and bumpy baseline. In contrast, the envelope spectrum generated by the FPT shown in panel (3) requires only 1700 signal points to converge and thereby clearly shows six peaks along the abscissa (chemi-cal shift in parts per million, ppm). From the heights of these peaks, the relative concentra-tions of several metabolites of interest can be assessed, including some of the components of total choline. This is clinically useful infor-mation, which the FPT algorithm provides at a short signal length, whereas the FFT needs signals more than 32 times longer (32 x 2048 = 65536) to generate a spectrum such as that shown on panel (3). The high-resolution of the FPT is due to its non-linearity and its interpola-tion and extrapolation capabilities, stemming from its form as a ratio of two polynomials P/Q, thereby facilitating noise cancellation [5, 6].

Given the superior resolution performance of the FPT compared to the FFT, we suggested how this resolution advantage could be applied to MRS data acquisition, by encoding fewer and relatively short MRS time signals [6]. A direct advantage of this high-resolution would be shorter examination time, with consequently lowered costs of the MRS procedure, not to mention increased comfort for the patient. Since volumetric coverage is often needed in breast diagnostics, the higher resolution of the FPT for MRS becomes even more vital for MRSI as spectroscopic imaging measurements together with 3-dimensional spatial resolution must be performed within a reasonable total acquisition time to be clinically practical.

AssessMeNt oF breAst cANcer MArker, pHospHocHoLINe VIA pAdé-optIMIzAtIoN

The converged envelope spectrum shown in panel (3) in Figure 1, would be the maxi-mum that the FFT could provide under ideal conditions. This is because the FFT is non-parametric and cannot give any information on what underlies the envelope. In panel (4) of Figure 1, it can be seen that phosphocholine (PC), the breast cancer marker, lies completely underneath phosphoethanolamine (PE). From panel (3) we would not even guess that the PC peak exists and its actual concentra-tion, which in breast cancer diagnostics is the key clinical question, would remain unknown if reliance was made only on the FFT.

The fast Padé transform is a parametric processor whose mathematical capabilities are well-suited to handling MRS data [5]. The component shape spectrum shown in panel (4) was generated from the signal parameters reconstructed by the FPT. From these parameters, the metabolite concentra-tions, including that of phosphocholine, are precisely computed as per panel (5).

sIGNAL-NoIse sepArAtIoN: crUcIAL For beNcHMArkING

A major barrier to more widespread appli-cation of MRS and MRSI in the diagnosis of breast and other cancers has been the pres-ence of noise within clinically-encoded MRS time signals. The FPT algorithm handles this via the concept known as “Signal-Noise Separation”. Especially with very closely-overlapping peaks, the number of genuine metabolites can be a very small percentage of the total number of resonances generated. In the case of breast cancer, there is a large number of spurious resonances with only seven true peaks within the windowed inter-val of 3.16 ppm to 3.34 ppm. These seven

true metabolites are clearly identified. Stable against noise, their full width at half maxi-mum (FWHMax) differ from full width at half minimum (FWHMin) and their ampli-tudes are non-zero. These are completely dis-tinguished from the spurious peaks which are unstable when noise level varies; whose FWHMax and FWHMin coincide [panels 6 & 7, Figure 1] and whose peak heights are zero [bottom right panel 8, Figure 1].

Overall, the use of more powerful math-ematics based on the fast Padé transform enables much more widespread applica-tion of MRS and MRSI techniques. The approach is particularly easy to apply since the basic techniques can be carried out using the same equipment (radiofrequency coils, amplifiers, etc.) as in standard MRI.

coNcLUsIoNMolecular imaging will inevitably have

an increasing effect in all areas of clini-cal decision-making in oncology. Recent advances in powerful and robust algorithms of proven validity show that the results of molecular imaging can become the basis of future personalized cancer medicine. Molec-ular imaging using magnetic resonance, in which signal processing methods are math-ematically optimized, has the potential to improve early cancer detection, which is par-ticularly useful in high risk patients.

AckNowLedGeMeNtsThis work was supported by cancerfonden, Radiumhemmets Research Fund and Stockholm city council (FoUU) to which the authors are grateful.

reFereNces 1. Kircher MF, hricak h and larson SM. Molecular imaging for

personalized cancer care. Molecular Oncol 2012; 6: 182.2. Belkić Dž, Belkić K. Molecular imaging in the framework

of personalized cancer medicine. Isr Med Assoc J 2013; 15: 665.

3. Barclay A. Over diagnosis: the new disease? Diagnostic Imaging Europe 2013; 28: 3.

4. von Schulthess GK et al. Integrated PET/cT: current appli-cations and future directions. Radiology 2006; 238: 405.

5. Belkić Dž & Belkić K. Signal Processing in Magnetic Resonance Spectroscopy with Biomedical Applications. Boca Raton, cRc Press Taylor & Francis Group, 2010.

6. Belkić Dž & Belkić K. Strategic steps for advanced molecu-lar imaging with magnetic resonance-based diagnostic modalities. Technol cancer Res Treat. December 2013, on line-first. DOI: 10.7785/tcrt.2012.500401

7. Begley JK et al. In vivo proton magnetic resonance spec-troscopy of breast cancer: a review of the literature. Breast cancer Res 2012; 14: 207-217.

8. Katz-Brull R et al. Metabolic markers of breast cancer. Enhanced choline metabolism and reduced choline-ether-phospholipid synthesis. cancer Res 2002; 62: 1966-1970.

9. Glunde K et al. MRS/MRSI guidance in molecular medi-cine: targeting choline and glucose metabolism. nMR Biomed. 2011; 24: 673-690.

10. Gribbestad IS et al. Metabolite composition in breast tumors examined by proton nuclear magnetic resonance spectroscopy. Anticancer Res 1999; 19: 1737-1746.

Figure 1: Clinically relevant data in Magnetic Resonance Spectroscopy (MRS) typical of breast cancer [10]. Input data: time signal: panel (1). Output data: spectra: panels (2), (3), (4); concentrations: panel (5); signal-noise separation in panels (6), (7) at FWHMax (full width at half maximum) equals FWHMin (full width at half minimum) for false data; true & false peak heights in panel (8). Symbols: input data (+), output data (O ,). To avoid clutter, FWHM doubled at 3.220 ppm (PC*) in panels (6), (7). Identification of false data by a 3-fold signature : (a) FWHMax = FWHMin in panels (6), (7) as (confluence of O & ), (b) zero peak heights in panel (8): abscissa crosses the centers of empty circles & (c) marked instability: panels (6), (7) for 2 noise levels differing by 10. Detection of true data by: (a’) FWHMax ≠ FWHMin, coincidence of O & + via ⊕, (b’) non-zero peak heights & (c’) stability against noise: panels (6), (7). Chemical shifts in parts per million (ppm), concentrations in µM/g, peak widths in 103 ppm & peak heights in arbitrary units (au). Noise level σ proportional to the root-mean-square (RMS) of the original time signal. Time signal, panel (1): Larmor frequency of 600 MHz, bandwidth (BW) of 6 MHz, its reciprocal: sampling rate τ. Absorption total shape spectra (envelopes) in FFT: panel (2), in FPT: panel (3). Absorp-tion component shape spectra retrieving PC underneath PE, in FPT: panel (4); impossible in FFT, post-processed by fitting. Concentrations, in FPT: panel (5). Signal length: total (N = 2048) in FFT & partial (N

P = 1700) in FPT. Key

clinical significance of parameters in MRS by FPT: peak heights in panel (8) multiplied by FWHM in panel (6) or panel (7) give metabolite concentrations in panel (5). With N = 2048, FFT in panel (2) yields only an envelope, lacks convergence, misses 3 metabolites, gives 4 broadened & severely shortened peaks of no clinical use. At N

P = 1700, FPT

converges providing the diagnostically critical information: concentrations in panel (5). To match FPT in panel (3) sampled at N = 2048 (2K), FFT needs N = 65536 (64K) signal points in a single encoding, causing a 32-fold lengthen-ing of each transient. In FPT, further improvement in signal-to-noise ratio (SNR) & total acquisition time is achieved using barely a quarter of the 128 transients needed for acceptable SNR in FFT. This is a twofold bonus for MRS via FPT: (i) higher accuracy: better clinical reliability, (ii) higher efficiency: shorter examination time of the patient.

FOCuSHospItAL


A vibrant health care center in eastern BelgiumSituated in the Belgium province of Limburg, close to the Dutch and German borders, the Ziekenhuis Oost-Limburg (ZOL) hospital is experiencing a steady increase in patients and, in the radiology department, there is a corresponding rise in the number of exami-nations being carried out.

The department has just undergone an extensive lay-out re-design and refurbishment to improve the patient experience.

On top of all this, the department has recently switched to a new PACS system.

We spoke to Dr Martijn Grieten, head of radiology, about this hive of activity.

Q Let’s start at the beginning. Please tell us about your hospital, the Ziekenhuis Oost-Limburg (ZOL)

Well, first of all I should say that ZOL is actually not one hospital but rather a group of three individual hospitals located in and around the city of Genk, which is one of the main towns of the Belgian province of Limburg. Lying on the borders of The Netherlands and Germany, Limburg is also not far from France and Luxembourg. Thus, within a 100 km radius around Genk, the so-called “Euregio” hinterland behind our hospital includes the Dutch cities of Maastricht and Eindhoven, the German city of Aachen and the Belgian cities of Antwerp, Brussels and Liège. So our hospital is a supra-regional public hospital, offering a complete care package from basic to highly specialized medicine. Over the three sites we have a total of 811 beds (and 256 daycare beds) with more than 2000 full time equivalent employees and no fewer than 236 physicians. Our particular expertise covers gyn/obs and infertility; oncology & radiotherapy; nephrology; neurosurgery; neo-natology; cardiology; stroke cardiac surgery; stomatology and intensive care.

Annually we have 35000 hospitalizations plus 67 000 “Day hospital” admissions and nearly half a million con-sultations/ambulatory visits. We have 45000 emergency admissions and 34000 surgical procedures carried out each year. There are 2000 babies born in the hospital each year.

However there is much, much more to our hospital than just a list of patient numbers and specialities. A core aspect of the overall philosophy of the hospital — and this

is not just an abstract concept, but one that we really try to put into practice — is patient friendliness. This emphasis on putting the patient at the centre of our activities is one that we also actively apply in the department of radiology.

Q Ok, let’s turn to radiology. What personnel and equipment are available to provide the service the

hospital expects? The overall mission of our department is clear: the pro-vision of safe and efficient medical imaging, using the appropriate imaging modality for the particular pathology involved or clinical question to be answered. In general we’re very well equipped with up-to-date imaging systems. Nevertheless in some areas such as MRI, we can be under particular pressure to provide the short waiting time and high quality service our referring physicians expect. We basically have all the imaging modalities: X-ray, mobile X-ray, CT, Cone beam CT, mammography units and biopsy facilities, interventional X-ray, MRI and ultrasound.

The main campus, Sint Jan, focusses more on the high-tech imaging modalities and the other campuses on more routine scans with high throughput and efficiency.

All told, as you might expect, the most common imag-ing procedure carried out is straightforward radiog-raphy, with more than 115000 X-rays performed each year. Next comes ultrasound with approximately 42 000 procedures per year. There are 38000 CT examinations and approximately 22000 MRI procedures each year. And, for two reasons this work-load is growing all the time. First there is the background aging of the population

Dr Martijn Grieten heads up a team of 20 radiologists, 63 nurses and technologists as well as 2 IT specialists, 16 medical management assistants, 2 research asso-ciates and a medical physicist. The ethos of the department is not just to produce high quality images, but also to avoid giv-ing the patients the feeling that they are simply an item on a production line, albeit a modern and high-tech one..


which brings with it more cases requiring radiological exami-nations. Then, the hospital itself is expanding so there are more doctors bringing with them more patient consultations.

On the personnel side, to support all this in our depart-ment we have no fewer than 20 radiologists & 2 assistants, 63 nurses and technologists, 2 IT nurses, 2 research associates, 16 medical management assistants, 4 logistics specialists and a medical physicist.

Q And on top of all this continuing, and growing, patient throughput, your department is just coming to the end of

a major refurbishment process. Actually it’s more than just a refurbishment it’s a complete

re-design of the layout of the department. This was started more than a year ago and is now nearing completion. And it is a real, practical example of what I was saying earlier, namely putting a priority on patient friendliness. So often radiology departments put the sole priority on the technical aspects of generating high quality images. Of course that is vital. How-ever the reality is that the patient, who is at the heart of all this, will probably never even see the images that we radiolo-gists are so proud of. The patient is much more likely to be impressed (or otherwise) by the actual circumstances of the imaging procedure and how he is received for his imaging exam. Apparently mundane aspects such as the condition of the waiting room, the facilities in the dressing rooms and the support of the staff during and after the exam are very impor-tant for the patient.

We made an active effort to address these issues, using expert consultants to advise us on the use of patient-friendly materials such as creative lighting, extensive use of wood panelling and, above all, providing spacious patient changing or preparation rooms. And since, in common with most hos-pitals, our department pays rent to the hospital on the overall space we use, an apparently simple decision such as providing more space for the patient changing or preparation room, is one we don’t take lightly. However since patient-friendliness is at the core of our values we considered the financial implica-tions of our decisions process to be worth it.

Gratifyingly, so far the reaction from our patients to the results of the refurbishment is very positive. Likewise, despite the inevitable inconvenience during the construction process, the staff appreciate the new working environment.

Q As if all this wasn’t enough, a year or so ago you decided to change your PACS system as well. Why was this?

We were talking earlier about the patient friendliness aspects of the refurbishment of our department. That’s impor-tant, but of course it goes without saying that behind this we need not only to generate high quality images (from all our modalities) but also have efficient PACS systems for our radi-ologists to produce the reports. These reports are, after all, the “end-product” of our department. The referring physicians will only consult the report rather than scroll through all the images. Efficient RIS/PACS systems and especially ones that can be optimally configured to meet our requirements are vital for the continuing optimal functioning of the department.

When it became clear that our previous old PACS system was coming to the end of its useful life, we knew that the continued efficient functioning of the department required a new PACS.

Q What system did you choose and why?We went for the Agfa Healthcare IMPAX system which

we installed about two years ago. In addition to the basic perfor-mance characteristics of the system, one of the reasons for this choice was that our RIS system was the Q.doc system from Agfa Healthcare. Efficient compatibility between RIS and PACS is essential; we considered that the likelihood of this is increased if both systems are provided by the same company. Although “pick and mix” approaches of RIS and PACS from different suppliers are theoretically possible, in practice issues such as compatibility of updates are much simpler when the same supplier is involved.

Q And how is the PACS performing in practice? Any teething troubles? What particular features do you appreciate most ?

We have the IMPAX 6.5.1 version and we’re very happy with the system. The installation process and transition from the old system itself was essentially trouble-free, taking a couple of months, mainly for the transfer of our 54TB memory of archived images. The close integration of the RIS/PACS is shown by the fact that the radiolo-gists are now mostly unaware that the RIS is actually running in the background when they’re reading their images. The stability of the system is fine with crashes almost non-existent. Our imaging systems are sourced from the main instrument suppliers, e.g. our CTs come from Siemens, Ultrasound is mostly from Philips and we have one GE MRI and one Siemens MRI system. Connectivity and PACS access to all these images is not an issue. Neither is the geo-graphical separation of the three hospitals that make up our group.

One minor inconvenience of the new system was the absence of a built-in 3D viewer. We solved this easily by incorporating the 3D viewer from the old PACS system we used to have. Apparently new versions of the Agfa Healthcare PACS system, will have an integrated 3D viewer, so we will be very interested in that.

In addition to the usual tools for the organizing of and pro-cessing of images provided in all PACS, there are two or three features of the IMPAX system that we particularly appreciate:

Work listsThe creation and organization of efficient worklists that we

can easily create with the new system is really useful. The custom-

In the redesign of the ZOL radiology department, particular attention was paid to patient-friendly aspects such as creative lighting, extensive use of wood panelling and the provision of large amounts of space in patient preparation rooms .


hOSpITAL FOCuS

ized configuration of the system is easy. For example, we have set up rules so that immediately the prioritization of patients can be visualized on screen via color codes indicating categories of patients such as emergencies; those from the intensive care unit or those on the waiting lists.

Structured reporting At the beginning, this feature was

viewed with a lot of scepticism, even resistance by many of our radiologists. There were several reasons for the ini-tial doubts, the most important being the fear that having pre-set fields in the report would in some way restrict the radiologist’s freedom and ability to correctly and accurately describe the images they were reading. In addition the radiologists had become very skilled at dictation and could easily and quickly incorporate all the necessary formatting notes into blank report forms.

However with the old non-structured system, the inconsistency of the final reports was recognized as a drawback. One radiologist might dictate a full page report for a case and another only a few lines for the same case.

What is very gratifying is that after the initial doubts, now that our radiologists have practical experience of the use of structured reporting, they are all converted and fully convinced of the advantages of it, even going to the extent of requesting that the approach be adopted as widely as possible.

Business Intelligence (BI) The IMPAX Business intelligence mod-ule (BI) is interconnected with the RIS and PACS. It is extremely useful for the

optimal management of the depart-ment since basically it allows unlimited mining of all our vast amounts of data. So, for example at a stroke I can see for each imaging modality a complete breakdown of our patients by age. Such information is vital for example to jus-tify the decision to make things easier for our more elderly patients by pro-viding more spacious preparation and changing rooms.

Another example is that the BI mod-ule enabled quantitation of the overload-ing of our MRI facilities that I alluded to earlier. This led to a reorganization so that we could extend our MRI activi-ties, which now operate seven days a week, 15 hours a day. Even with this, our MRI waiting lists are continuing to grow. Again data from the BI module showed that, at least in part, this was due to some MRI patients coming across the border from The Netherlands, where initially waiting lists were even worse than ours. Clearly a concerted approach to the question of MRI supply and demand is needed. The BI mined data is indis-pensable in deciding exactly how the problem should best be tackled.

At a much more immediate level, the BI module also provides dashboards which can show key performace indi-cators that are continually monitored. For example, the progress of patients through the imaging process can be seen in realtime, quantitating the numbers of waiting patients or those awaiting results, etc. In this way, bottlenecks can be highlighted or regular time ineffi-ciencies in patient workflow identified so that, more importantly, timely correc-tive action can be initiated.

Quality Assurance Audit for Diagnostic Radiolo-gy Improvement and Learning

(QUAADRIL)

Under the leadership of Dr Mar-tijn Grieten, the radiology department of the ZOL hospital in eastern Bel-gium was the first in the European Union to volunteer for an indepen-dent quality audit of its activities. This was carried out under the aegis of the International Atomic Energy Agency (IAEA). Initially focussed on quality aspects such as dosimetry and calibra-tion, IAEA has extended its activities to the whole field of diagnostic radiol-ogy and has introduced the Quality Assurance Audit for Diagnostic Radi-ology Improvement and Learning (QUAADRIL) program.

The increasing interest in QA and quality improvement in diagnostic radiology is driven by a number of factors. These include the high cost and increasing technical complexity of radiological equipment, the awareness of the possibility of increasing doses to patients and the general importance of radiological diagnosis to patient management.

In practice the audit process involved the evaluation by an inde-pendent team of external experts (radiologists and medical physicists) of the data, documents and resources to check performance against stan-dards. Lasting just under one week, the audit is essentially a process of fact finding and interpretation. The inde-pendence of the external auditors has the advantage of giving a broader and more universal perspective. In addi-tion the drawback of internal depart-ment audits is avoided, namely when internal auditors may be so accus-tomed to long-standing and routine practices that they cannot identify any weaknesses or limitations of their own operations.

Described by Dr Grieten as inter-esting, rewarding, but intensive, the QUAADRIL audit was completed mid June this year. The radiologists of ZOL await the formal notification of the audit results with quiet confidence.

As part of the refurbishment of the radiology department at ZOL, new radiology reading stations were installed which are insulated for sound and have the possibility of personalisation of background light levels. Here Dr Monique Horvath, principal radiologist at ZOL uses the “chat” function of the IMPAX system to discuss details of images with colleagues.

d6157_Pub_EE2014_A4.indd 1 23/01/2014 09:39

CANCeRbreAst


By Dr B. Scheffer

ultrasound elastography in Screening and Diagnosis of Breast CancerA systematic program of breast ultrasound screening was carried out in our clinic over a period of more than six years on a large number of women presenting for mammography examination. The aim was to analyze the ability of ultrasound elastography to reduce the number of unnecessary biopsies and to evaluate the overall potential of the technique. Out of more than 16000 patients examined by ultrasound (HI-RTE, Hitachi Medical), 271 breast cancers were detected, 27% of which had not been detected in mammography. An additional examination using ShearWave Elastography (SWE, Supersonic) was performed on 100 lesions found to be neoplastic. The main contribution of real-time elastography was the increase in confidence that biopsies were not being taken unnecessarily: the use of RTE gave a positive predictive value (PPV) of 75% for a positive biopsy result. The PPV can be increased even further, up to 82%, through use of a secondary, comple-mentary examination using SWE. This article summarizes our experience of the use of the techniques, their indications and limitations.

Two factors have been cited for the welcome reduction in the global mortality rate from breast cancer, namely

- the increased detection of smaller lesions through the use of screening mammography

- progress in neoadjuvant therapies and the improved con-trol of resection boundaries

In the context of the increased detection of smaller lesions, studies have shown that B-mode ultrasound of breast nodules has a high negative predictive value (NPV) but only a very modest PPV. Despite the clear advantages of the technique such as easy accessibility and lack of ionizing radiation, ultra-sound breast screening has the problem that can be summa-rized as enabling the detection of a high number of lesions and abnormalities that do not require biopsy because only very few are neoplastic.

For 6 years we have conducted a systematic ultrasound screening exam to back up no fewer than 16107 mammo-grams. In our study we combined elastography for each abnor-mality revealed in B-mode.

MetHodsPost-mammography ultrasound screening was performed using a Hitachi scanner (EUB Elite HI-Vision). The Hitachi Realtime Elastography (HI-RTE) strain elastography mod-ule for this system displays the relative difference in elasticity of breast tissues when subjected to an external mechanical compression initiated by the operator. The simultaneous dis-play of the B-mode image, the elastogram and the scale or compression curve allows the operator to assess the quality of the technique and to check that the lesion remains in the field

of examination during the data acquisition. By convention, a color scale is used to present the results. Suspicious, stiffer lesions are coded blue.

In the latter part of our study all suspicious lesions imaged with HI-RTE were additionally analyzed using the Aixplorer ShearWave Elastography (SWE) system (Supersonic Imagine). With this system the results are shown in kilo pascals (kPa). In our study the same conventional color coding of blue for stiff lesions was adopted.

The standard examination procedure with the Hitachi system consisted of a sweep of both breasts, quadrant by quadrant, together with a study of the axillary region. Each B-mode abnormality, nodule or region of disorganised ultra-sound architecture, was then examined using HI-RTE in two orthogonal planes with two transducers.

Examinations using SWE were also performed using two orthogonal planes but, for patients whose lesions were not located at the junction of two quadrants, radial and anti-radial sections were also carried out. SWE was performed using a kPa scale from 0 to 80. The use of this scale gives greater sensitivity than the scale of 0 to 180 kPa which is used in many published studies on SWE.

LeArNING process Because, elastography, like ultrasonography itself, is operator-dependent, we set up a rigourous program of operator auto-evaluation. Validation of the success of the learning process of probe movement and interpretation never took more than 6 months. Nevertheless, for the first year the results of the HI-RTE elastography were not used to alter the clinical management as determined by the B-mode result. After this initial period, patients with lesions that were elastographically benign were put on a simple ultrasound follow-up schedule, with exams at 6 months, 12 months and then at 24 months. During the evalu-ation period results obtained with the SWE technique did not change the clinical management determined by HI-RTE .

the author

dr b. scheffer is at

centre Radiologique ZOlA, nantes, France

email: [email protected]


All lesions that were suspicious on HI-RTE were verified his-tologically; only a few benign lesions on HI-RTE were evaluated with SWE.

Mammography and ultrasound examinations were all per-formed by a single radiologist who also carried out the initial clinical examination of the patient. The description of all radio-graphically occult lesions was validated by an independent radiol-ogy reference site, who were unaware of the results of the clinical examinations. Thus, the mammographic analysis carried out in the reference centre was not influenced by the clinical exam.

INterpretAtIoN oF eLAstoGrApHY resULts. HI-RTEWe classified nodules according to the 6-level elastography scoring system described by Hitachi, using the conventional color coding

representation of tissue stiffness [Figure 1a, b]. The Fat-Lesion ratio (FLR) is a semi-quantitative measure of the ratio of elasticity of two areas (fatty tissue and a lesion) and has been proposed to dif-ferentiate malignant and benign lesions. Unfortunately in practice FLR has several problems such as possible lack of homogeneity of the reference fatty tissue. Also comparison of the published studies FLR is difficult since they use different cut-offs. In future the Gel to Lesion ratio (GLR) method may overcome the deficiencies of the FLR method. In GLR, a pad covering the probe contains gel of a known, standard elasticity so could serve as a reference value.

The decision-tree we used for the incorporation of HI-RTE into the diagnostic work-flow is shown in Figure 2.

Shearwave ElastographyWe used a 3-score classification in SWE, namely scores SC1, SC3 and SC5 [Figure 3]. These were determined on the basis of three parameters, namely the color map, the elasticity index and the lesion/fat elasticity ratio. The score (i.e. SC1, SC3 or SC5) would normally be used to choose the precise clinical management options for the patient. However, during our evaluation period, these scores were only used for information purposes. The results were not used to actually decide clinical management.

coNtrA-INdIcAtIoNs ANd precAUtIoNsReal Time Elastography and Shear Wave Elastography are both used for the determination of the hardness/elasticity of a lesion, but they involve two fundamentally different technological principles. They each have specific advantages and limitations.

However both techniques share certain absolute contra-indica-tions. One of these is the presence of local hemorrhaging, which can alter the stiffness of the lesion. Elastography examinations should therefore always be performed prior to any invasive procedures.

In addition, in both techniques, particular care should be taken in certain cases:

•When there is spontaneous or acquired hardness of the breast, such as in cases of major fibrous mastopathy or an inflam-matory or carcinomatous mastitis, the hardness of the breast can give rise to artefactual interpretation of HI-RTE images [Figure 4]. Elastography should not be used in such cases. With SWE we observed an elevation in the elasticity index in the surrounding tis-sue with less homogeneous elasticity mapping in these cases.

• It has been shown that there is a benefit in using elastogra-

FigUre 1. Based on the conventional color-coding representation of the elastographic stiffness of tissue (soft is green, hard is blue) the scoring system (Fig 1a) used in our HI-RTE studies comprised six scores , 1-5 plus a score RVB ( Red Green Blue). Score 1 nodules are is entirely elastic In Score 2 the majority of the nodule is elastic with some small stiffer areas. In score 3, the center of the nodule is hard, but the periphery remains elastic. In Score 4 the entire nodule is hard. Score 5 nodules are represented as blue and surrounded by a blue halo. In “RVB” the superimposition of the three colors red, green, blue (RGB) corresponds to a coding artefact in cystic formations. In addition to these scores we used a “Mosaic” Score (Fig1b) for non-mass lesions that are more or less echogenic with a disorganization of the ultrasound architecture. Lesions with score 3, 4, 5 or mosaic were considered as indications for biopsy.

Histology

malignantFollow-up

Treatment

Stop Benign Risk factors

Learning Process ELASTOGRAPHY

Score 3 4 5 Mosaic

Unstable signal or negative but presence

of microcalci�cations

Score 1RGB

Score 2

B mode strategy

Malignant signs? Benign signs?Management as a

function ofBI-RADS class

Known lesionor BI-RADS 2

FigUre 2. In conventional ultrasound, the usual diagnostic strategy is firstly to eliminate know stable lesions and also BI-RADS 2 lesions and then to look for signs of malignancy. The presence of benign signs will result in a classification of the lesion as BI-RADS 3 with follow-up. The absence of malignant or benign signs leads to a classification of the lesion as BI-RADS 4a with the need for further investigation, namely biopsy. During our elastography learning curve, all lesions were analyzed to give a better under-standing of the technique and semiology. In routine practice, whatever the BI-RADS classification in B-mode, the lesion was analyzed using elastography. Only lesions showing signs of instability and microcalcifications visible in mammography without an identified ultrasound mass passed again to the conventional diagnostic path.


BReAST CANCeR

phy to distinguish malignancy associated with microcalcifications detected on mammography, in particular to determine the risk of invasion in ductal carcinoma in situ (DCIS). In practice, when the mammary gland volume is large it can be difficult to find the zone of microcalcification.

• In the absence of calcified cystosteatonecrosis, a scar after simple surgery is attributed a score of 1; after conservative radio-surgical treatment, preferentially a score of 3, sometimes a score of 2, rarely a score 4 and never a score 5.

•Tumor necrosis induced by chemotherapy modifies the elas-tography mapping and gives a ‘soft’ pattern. Chemotherapy is gen-erally a contraindication for elastography.

• Radiotherapy. Major post-radiation cutano-glandular sclero-sis is rare but can be a limiting factor for elastography. Usually elas-tography can be performed after radiotherapy without any particu-lar restrictions, although a small loss of elasticity may be noticed.

• Breast Implants. The presence of an implant is not a strict contra-indication for elastography but could in theory pose a prob-lem if, in a small breast, the lesion is situated in contact with a tight implant, which could artificically increase the hardness of the breast. In our study the examination of several benign lesions in contact with an implant were carried out without difficulty. The only malignant lesion that we examined by HI-RTE in patients with implants was not in contact with the implant.

• Influence of the menstrual cycle. It has been shown that the HI-RTE color map does not change from one menstrual cycle to another. Without oral contraception, the stiffness of the paren-chyma was significantly increased during the luteal phase compared to the menstrual period and follicular stage. Elastography should therefore be more discriminating in the first part of the cycle. In practice it is necessary to perform follow up elastography examina-tions at the same stage of the menstrual cycle.

FActors AFFectING resULts. Depth of lesion This is one of the principal factors affecting the quality of elastogra-phy results. In HI-RTE the use of a waterbag stand-off allows easy examination of superficial lesions. In SWE it might be possible to obtain similar results by using a stand-off kit to suppress the surface artefacts created by reverberations from the skin surface. For deep-seated lesions, it is necessary in HI-RTE to lower the frequency of the transducer. In SWE a SL10-2 transducer should be used (not tested in our study) which theoretically should give identical results.

Lesion Size With HI-RTE, the elastography signal of lesions of approximately 5mm diameter is independent of their surroundings. With SWE, the elastography map shows hardly any heterogeneity for small lesions essentially surrounded by fat. However this is not the case for lesions of the same size within the gland. In a fat environment, the same phenomenon is observed for lesions of 10 mm situated deeper. As for lesions greater than 20 mm, the HI-RTE signal in the centre of such lesions may be heterogeneous, and more or less ‘soft’ because of necrosis, but the periphery remains hard. It is sometimes necessary to take only part of the lesion in the Region Of Interest (ROI) in order to respect the general guideline whereby the ratio of the lesion to other tissue should not exceed 25%. With SWE central necrosis is also found in such lesions.

resULtsOf the 271 malignant lesions which were analyzed using the HI-RTE system, 100 were also analyzed using the SWE technique.

b-Mode screeNING coMbINed wItH eLAstoGrApHY A bi- or multi-focal lesion was found in 30 patients and a bilateral lesion was found in eight patients. Of these patients with multifocal lesions, B-mode screening missed five lesions, all of which had been detected by other modalities; one case had been detected during the clinical examination carried out by the oncologist and the other four cases were detected by MRI.

A total of 271 malignant lesions were analyzed. Of these lesions, 27% had not been detected on mammography, and of these 41% were invasive lobular cancers (ILC), the majority being grade II and

FigUre 3. In Shear Wave Elastography, a 3 -score classification system is used: SC-1, SC-3 and SC-5 . These scores are calculated from three parameters.: 1) The color map. In this qualtitative parameter, homogeneously mapped lesions are classified as score 1 and heterogenous lesions are scored as 3. A score of 5 is attributed when the lesion and the peri-lesional area are heterogenous. 2) The elasticity index, which is measured at the level of the hardest zone as indicated by the color map. The cut-offs used are 15, 30 and 70 kPa for scores 1, 3 and 5 respectively. 3) A semi-quantitative parameter: the lesion/fat elasticity ratioThe principle is the same as that in HI-RTE and has the same limitations. The LFR ratios used as cut-offs were 1, 3 and 7 for scores 1, 3 and 5, respectively. Note: In the diagram above “Moy” stands for average, “Ecart Type” for standard deviation and “Rapport” for ratio.

FigUre 4. IDC 10mm a)left breast with a stony hardness, density 4 on mammog-raphy b) B-mode ultrasound, hypoechogenic nodule with a horizontal axis c) Hi-RTE weak mosaic d) SWE positive (EI=49 kPa) e)MRI.


III. The mean size of such lesions was 8 mm compared to an aver-age size of 14 mm for lesions that were detected radiologically. The distribution of histologically confirmed invasive ductal carcinoma (IDC) was similar to that reported in the literature, in particular, for carcinomas with colloid and lymphoid differentiated forms. These are theoretically soft cancers and are therefore not mis-identified by elastography screening methods. The variant Score 4 with some small soft spots is predominantly found in these histological varieties.

reprodUcIbILItY: General case With HI-RTE, the images were found to be reproducible over several sequences, independently of the size of the lesion, with two transducers and in two planes [Figure 5]. B-mode abnormali-ties were only determined positive with elastography under these conditions. The HI-RTE technique analyses the deformation of a volume relative to surrounding tissues and a tumor volume must be detected in two orthogonal planes. The color map reproduces the shape of the lesion in B-mode.

With SWE, several acquisitions were performed for each lesion. The color map was less reproducible because this technique analyses the hardness of the lesional and peri-lesional tissues point-by-point and it is difficult to obtain exactly the same section in the same plane [Figure 5]. Lesions are not histologically homogeneous and the elasticity index and therefore the color map are necessarily different in each plane, without necessarily changing the classification of the lesion. In two cases with small lesions, the SWE color map varied from positive to negative depending on whether the section was radial or anti-radial. The practical consequence of this is that it is necessary to examine all small lesions at their orthogonal incidence.

reprodUcIbILItY: case of unstability or non-reproducible elas-tography signals In some cases the signal obtained with both techniques was not stable. Such instability could have either a single cause, or several causes that are more or less inter-connected, e.g. heterogeneous texture of the breast, deep or superficial location of the lesion or particular benign or malignant histology. In all cases when the elastography signal was unstable, we did not use the result but instead used the conventional B-mode images for diagnosis and future management decisions.

reprodUcIbILItY: interpretationPreviously published HI-RTE studies from other groups and

involving a total of 1668 cases reported good inter-observer repro-ducibility, with a kappa score ranging from 0.73 to 0.90. In another study, involving only 63 cases, kappa was 0.59.

As for SWE, one study involving 758 cases, reported kappa scores of 0.57 and 0.66 for the color map and maximum elasticity index measurement respectively.

These elastography results contrast with the reproducibility for the classification of BIRADS 2, 3, 4a, 4b, 5, and overall classification (κ = 0.27, 0.32, 0.14, 0.16, 0.26 and 0.28 respectively).

tHe doUbLe VALUe oF eLAstoGrApHY

screeNING: INcreAsING tHe ppV oF tHe MIcrobIopsYThe HI-RTE technique is reliable for identifying lesions that require microbiopsy. In our study the longitudinal follow-up over 6 years was carried according to a decisional tree [Figure 2]. This did not take into account the BIRADS classification as determined by B-mode except in cases where its use was restricted as described

above in the section on contraindications and precautions. No lesions that were elastographically benign on HI-RTE were subse-quently found to be malignant on follow-up.

Only a few lesions that were benign on HI-RTE were also exam-ined by SWE. Our study did not evaluate the discriminatory char-acteristics of the SWE technique without pre filtering by HI-RTE. During the period 2011-2012, 5699 mammography patients were examined using HI-RTE and SWE techniques.

HI-RTE elastography was positive for 133 lesions, of which 100 were cancers, 24 of them not detected by mammography. The SWE result was in agreement with HI-RTE in 86 cases (score 5); negative (score 1) in 10 cases for lesions that were either very superficial or entirely surrounded by fat; intermediate (score 3) in four cases, of which two lesions had different signals depending on the orienta-tion of the scanning plane. In our series, 33 benign lesions were positive on elastography and biopsied.

For the four adenofibromas biopsied: -one lesion, BIRADS 4a in B-mode, presented with an unstable

elastography image, negative on SWE- two BIRADS 4a lesions, weak mosaic pattern on elastography

but hypervascular in Doppler mode and positive with SWE - one lesion BIRADS 3 on mammography, score 3 with HI-RTE,

negative with SWE. The other benign lesions were 12 papillomas, 7 radial scars, 5

fibrous mastopathies (3 negative with SWE), 2 sclerosing adenosis, 2 cases of pseudoangiomatous stromal hyperplasia PASH (one negative with SWE), one lymphocytic lobulitis. The papillomas and radial scars were diagnosed without elastography either from the

FigUre 5. Reproducibility of images. a) HI-RTE. A 4mm invasive ductal carcinoma (IDC). Orthogomnal sections. L53 is a 10 MHz probe and L65 is a 14 MHz probe. b) SWE. A 5 mm invasive ductal carcinoma (IDC). Different mapping on two slighly displaced sections. c) SWE. A 4 mm invasive ductal carcinoma (IDC). Mapping either positive or negative depending on whether the section is radial or anti-radial.d) HI-RTE: Unstable signal, fibrous mastopathy.


BReAST CANCeRmammogram or via B-mode ultrasound. Four papillomas and one radial scar were negative on SWE. Because of the risk of papillary cancer for the papillomas and tubular carcinoma for the radial scars, it is necessary to verify their histology via sur-gical removal of the lesion. If they are not considered as false positives, the PPV of the microbiopsy indicated by HI-RTE elas-tography is 89% and 97% when combined with SWE. If these lesions were included as false positives, the PPV of the microbiopsy would be 75% and 82% when combined with SWE.

Thus by excluding the data for the papil-lomas and radial scars which require sur-gery, and which we consider as ‘useful false positives’, the addition of ultrasound screen-ing to the mammography screening gener-ated six “true” false positives out of 5263 patients examined (0,011‰). If the SWE results had also been taken into account, 3 lesions would not have been biopsied because their color map was perfectly homogeneous even with a scale adjusted to 0 - 40 kPa, and so the level of false positives would have only been 0,006‰.

dIAGNostIc: tHe seArcH For MULtI-FocALItY ANd stUdY oF LYMpH Nodes Elastography has played a vital role in the assessment of multifocality in the search for small satellite lesions which are not seen on mammography and only poorly identified in B-mode. In our study, 30 patients pre-sented with bi- or multi-focal lesions, and eight with bilateral lesions. The usefulness of intra-operative studies of the sentinel lymph node, either using conventional his-tology or the One Step Nucleic acid Ampli-fication (OSNA) test is not put in question by elastography results, but pre-operative suspicion of infiltration of the lymph nodes can modify the therapeutic approaches. A previously published study reported the exploration of 64 axillary lymph nodes (histology: 33 benign and 31 malignant) using HI-RTE and found a score varying between 1 and 4 as a function of the elas-ticity percentage of the node. The accuracy was 73.4%, with a sensitivity of 80.7% and of 87.1% when combined with the B-mode findings. In 81 lymph nodes (histology: 70 benign and 11 malignant) analyzed by SWE, a significant difference (p<0.05) was found in the measured value of elasticity of metastatic nodes, with an area under the Receiver Operating Curve (ROC) of 0.76 and 0.75 respectively for the mean and maximum elasticity index.

In practice, the diagnostic performance of the elastographic analysis of lymph nodes depends on the morphology of the patient and the depth and size of the targeted node. For deep adenopathies in patients whose axillar anatomy doesn’t allow the easy car-rying out of the compression-decompres-sion movements, the elastography signal is unstable, non-reproducible and therefore of low, or zero, diagnostic value.

coNcLUsIoNs Various studies have shown that screen-ing with ultrasound can make a useful diagnostic contribution but have also highlighted the problem of an increased number of false positives, which not only cause stress for the patient but also increase the overall financial cost. The addition of elastography is an effective counterbalance to the high level of false positives since, when used as an initial test the HI-RTE technique has a high PPV.

The “unhelpful false positives” generated by HI-RTE are essentially fibrous lesions, caused principally by a modification of intercellular substances. Such variations give rise to sclerosis in the case of collagen and elastogenic sclerosis in the case of elastin. Sclerosing collagen is hard, and elastogenic sclerosis is soft; their relative proportions can vary. In HI-RTE, volume analysis does not allow differentiation of these processes, so overall the zone will appear hard. With SWE, the point by point analysis has the advantage of defining the relative hardness of the zone. In our study all the lesions that were fibrous on histology appeared hard with HI-RTE but some were soft with SWE. If histology using a trichromic stain with saffron confirms this hypothesis, the use of the two modalities functioning indepen-dently in the same equipment could be par-ticularly interesting.

With SWE we came across two prob-lems: lesions surrounded by fatty tissue and lesions which give a different signal depending on the orientation of the ultra-sound plane. For the former, i.e. lesions situated in a fatty environment, it didn’t matter whether the lesions were small and superficial or larger and deeper: only barely visible on the color map with a scale set at 0-80 kPa, they are effectively invisible on a scale set at 0 to 180 kPa. The maximum elasticity index was very low, below our theoretical threshold limit of 30 kPa for score 3. Histologically, these cases were infiltrating lobular or ductal

carcinomas without particular differen-tiation or elevated grade. The explanation for this phenomenon may be a failure to focus the ultrasound or a different stromal reaction in a fatty environment. With SWE, certain small lesions pro-duced a different color map depending on whether the plane was radial or anti-radial. This phenomenon of directional dependency, or anisotropy, has not been seen in SWE examinations of organs such as liver or thyroid but has been reported for muscles. Elasto-histological studies of oriented operating room specimens will perhaps determine whether this theory of anisotropy is applicable to the breast whose histological architecture is in a radial orientation.

The BE1 study published in 2012 is considered as a reference for the SWE system in breast cancer but was not a screening study. The lesions in the BE1 patients had previously been detected by palpation, mammography, MRI or a prior ultrasound examination. In addition, sev-eral exclusion criteria were applied in the recruitment of BE1 patients using infor-mation gained from previous examina-tions, which of course would not apply in a screening situation. The use of the SWE technique for screening thus requires fur-ther clinical and histological studies.

FUtUre perspectIVesThe work we have carried out in the large number of patients studied so far clearly shows the value of HI-RTE elastography in decreasing the chance of unnecessary biopsies.

Elastography will continue to play a vital part in patient diagnosis and screen-ing in our clinic for the foreseeable future.

HI-RTE will be used as a first option, complemented by SWE in particular cases, for example fibrous lesions.

The future development of a single instrument incorporating both elastog-raphy modalities would of course sim-plify the practical implementation of this strategy.

FUrtHer reAdING There is a large published literature in the field. A complete bibliography is available from the author. 1. hitachi Real-time Tissue Elastography: Publications

& International communications clinical Abstracts. Accessible on http://tinyurl.com/hI-RTE-bibliography

2. Berg WA, cosgrove DO, Doré cJ, Schäfer FK, Svensson WE et al. Shear-wave Elastography Improves the Specificity of Breast US: The BE1 Multinational Study of 939 Masses. Radiology February 2012; 262: 435.

NeWSINdUstrY


cardiac contra-indications removed for ultrasound con-trast agent

An extensive evaluation has been car-ried out by the Committee on Human Medicinal Products (CHMP) of the Euro-pean Medicines Agency (EMA) of the ben-efits and risks of SonoVue, the ultrasound contrast agent from Bracco Imaging. Fol-lowing the review, the decision was made to remove the contraindication for use of SonoVue in patients with recent acute coronary syndrome or clinically unstable ischemic cardiac disease. Therefore, the product can now be used, even if with extreme caution, in patients with condi-tions such as evolving or ongoing myocar-dial infarction, typical angina at rest within last 7 days and significant worsening of cardiac symptoms within last 7 days. Other indications include recent coronary artery intervention or other factors suggesting clinical instability (for example, recent deterioration of ECG, laboratory or clini-cal findings), acute cardiac failure, Class III/IV cardiac failure, or severe rhythm dis-orders. In view of the frequent use of the direct-acting inotropic agent dobutamine for stress echocardiographic procedures, the CHMP also recommended to avoid the use of SonoVue in combination with dobutamine in patients with conditions suggesting cardiovascular instability where dobutamine is contraindicated.BRAccO IMAGInGMIlAnO, ITAlYwww.braccoimaging.com

FdA approves Gd contrast media for breast MrI

The U.S. (FDA) has just approved a new indication for Bayer’s gadolinium-based contrast agent Gadavist (gadobu-trol) for intravenous use with MRI of the breast to assess the presence and extent of malignant breast disease. The

approval was based on priority review of two, multi-center, Phase 3 studies (GEMMA 1 and 2) conducted in 13 countries. “The Phase 3 GEMMA stud-ies demonstrate that Gadavist-enhanced breast MRI provided a statistically sig-nificant improvement in the identifica-tion of the extent of breast cancer versus unenhanced MRI,” said GEMMA PI Dr Gillian Newstead, of the Univer-sity of Chicago Medical Center. In the GEMMA studies a total of 787 patients with recently diagnosed breast cancer from 13 countries were enrolled. MRI images were analyzed by three indepen-dent radiologists. These readers con-firmed that Gadavist-enhanced breast MRI improved ability to assess the pres-ence and extent of breast cancer when compared to images from unenhanced breast MRI. A key finding from the studies was a superior sensitivity for the presence and extent of malignant disease

compared to unenhanced breast MRI for all six readers. A true-negative rate (specificity) of >80% for breasts with-out malignant disease was confirmed for Gadavist-enhanced breast MRI by 5 of 6 readers. The studies did reveal that in breasts with malignancy, Gadavist-enhanced breast MRI overestimated the histologically confirmed extent of malignancy in the diseased breast in up to 50% of the patients. BAYER hEAlThcARE BERlIn, GERMAnYhttp://bayerimaging.com

Medical Imaging at the world cup A complete medical imaging depart-

ment has been created at the Argentin-ian training base in the 2014 World Cup. The centre includes a dedicated MRI system at the training centre itself, and extends to ultrasound scan-ning of injuries ‘on-pitch’ if required.

Esaote, the Italian-based manufac-turer of medical diagnostic systems, has helped the Argentinian team’s medical staff to establish the facility. The centre is equipped with Esaote’s O-Scan, a small and highly efficient MRI system specifically designed for imaging extremities such as knees and ankles, and the company’s fully portable MyLab ultrasound system. The Argentinian team doctor, Dr Ale-jandro Rolon, said “On the pitch I am able to diagnose not only the musco-lotendinous pathologies but also many other bone and joint related injuries. Whilst ultrasound is the main tool, the O-Scan MRI back at the camp will con-firm the diagnosis and help optimize treatment.”ESAOTE GEnOA, ITAlYwww.esaote.com

Marketing agreement for MrI post processing package

GE Healthcare and CorTechs Labs have just announced the signing of a strategic joint-marketing agreement, whereby the companies will collaborate to co-market powerful MRI scanning solutions com-bined with NeuroQuant, MRI post-pro-cessing application from CorTechs Labs. NeuroQuant enables the acquisition of volumetric brain image data and auto-matically conducts accurate and consistent measurements of cortical and subcortical volumes targeted at identifying evidence of neurodegeneration. This provides neu-rologists and radiologists with objective support for clinical analysis. “Accurate

INDuSTRY NeWS


measurement of volumes of different parts of the brain may be a useful aid in the diagnosis of neurodegenerative diseases such as Alzheimer’s, multiple sclerosis and traumatic brain injury,” says Chris Airriess, CTO of , CorTechs Labs. “In the literature volume changes in certain brain structures have been linked to early indications of such conditions where more traditional diagnoses lack detection sensitivity.” Neu-roQuant does this in an automated fashion and with no user intervention,cORTEchS lABS, SAn DIEGO, cA, USAwww.cortechslabs.com

French ultrasound manufacturer makes strong impact in Israel

SuperSonic Imagine, the French company specializing in ultrasound medical imaging, has announced the strong performance of its Aixplorer ultrasound system in the Israeli med-ical imaging market. The company has been marketing Aixplorer, its new generation ultrasound platform, in Israel since 2011 via its exclusive dis-tributor, Inframed Ldt based in Lod, Israel There are now approximately twenty Aixplorer systems currently installed in various medical cen-ters such as the Laniado Hospital in Netanya, the Ichilov Medical Center, the Assaf Harofeh Medical Center, the Edith Wolfson Medical Center and the Sheba Medical Center in Tel Aviv, the Hala Breast Clinic and the Hadassah Medical Center in Jerusalem,and the Barzilai Medical Center in Ashekelon. The Assuta Hospital in Tel Aviv has also recently ordered three Aixplorer ultrasound systems, in addition to the two already in use in the hospi-tal’s radiology ward. Pascal Dardelin, Supersonic Imagine’s sales manager for the Middle-East, explained, “Aix-plorer has been extremely successful among the Israeli scientific commu-

nity thanks to its to its cutting edge innovation, clinical benefits and ease of use. In the search for clinical reli-ability, Israeli doctors are always open to new technological innovations.” Aixplorer is the only ultrasound sys-tem that uses real-time ShearWave Elastography and UltraFast Doppler. It is a new generation system that provides exceptional imaging qual-ity thus helping physicians to better characterize lesions through the visu-alization and non-invasive measure-ment of tissue stiffness.SUPERSOnIc IMAGInEAIx-En PROVEncE, FRAncE www. supersonicimagine.com

Agreement on supply of Molybenum 99 to Us medical imaging market

NorthStar Medical Radioisotopes LLC has signed an agreement with Triad Isotopes Inc to establish the first commercially viable US source of molybdenum-99 (Mo-99), the most widely used radioisotope in medical diagnostic imaging. Under the agreement, NorthStar will work with Triad Isotopes to bring non-uranium-based Mo-99 produced with its proprietary RadioGenix iso-tope separation system to market. NorthStar, is developing a US source of Mo-99 produced without the use of highly enriched uranium (HEU) to help alleviate chronic shortages of the vital medical isotope. Mo-99 is the parent isotope of technetium-99m (Tc-99m), which is used in mil-lions of nuclear medicine procedures annually. Key applications include myocardial perfusion imaging, sen-tinel node mapping, hepatobiliary imaging and inflammation and infection imaging. Currently, almost

all Mo-99 is produced using highly enriched uranium at aging facilities located outside of the United States, leading to product shortages. North-Star has two methods of producing Mo-99 without using highly enriched uranium, both of which produce only a benign waste stream. nORThSTAR MEDIcAl RADIOISOTOPES llcMADISOn, WI, USA www.northstarnm.com

philips to supply new karolinska University Hospital

Philips has announced that it has signed a partnership agreement with Stockholm County Council (SCC) for R&D, an innovation program and education, and the procurement, installation, maintenance, upgrad-

ing and replacement of most of the medical imaging equipment for the new hospital site for Karolinska in Solna, Sweden. The contract was awarded following a public European tendering process and gives the SCC an option to extend the contract by another six years.

New university hospital facili-ties in Solna, part of the metropoli-tan area of Stockholm, are currently under construction and scheduled to be opened in late 2016. Today, the existing Karolinska University Hos-pital is one of the leading academic hospitals in the world and the new hospital has the ambition to further strengthen this position. The new Karolinska University Hospital will take a truly patient-centric approach to health care, and develop new care pathways by integrating patient care, clinical research and education. As a major teaching hospital, the new hos-pital will help to educate the medical specialists and healthcare profession-als of the future.


In line with this approach, the agreement has a strong focus on inno-vation and collaboration. The new Karolinska University Hospital will have access to state-of-the art imag-ing solutions and services, including those of other vendors, for a multi-year term at predictable costs. The managed equipment services include the procurement, system integration and timely updates to ensure that the new hospital will be provided with the imaging functionalities that its departments require for the optimal delivery of care. Moreover, Philips will establish a research & innovation hub at the new Karolinska University Hospital that will bring researchers from the medical technology indus-try, hospital and academia together to help facilitate a closer link between the delivery of care and clinical research. This is expected to contrib-ute to new research findings being translated into new therapeutics and treatment methods faster

The procured imaging equipment such as MRI, CT, ultrasound and inter-ventional X-ray systems will be used throughout the new hospital in radiol-ogy, cardiology, neurology, oncology, and pediatrics for both patient care and clinical researchPhIlIPS hEAlThcARE, EInDhOVEn, ThE nEThERlAnDS www.healthcare.philips.com/

Uk call for regulation of over-seas teleradiology

The UK Royal College of Radi-ologists (RCR) is calling for changes to be made in the way that over-seas doctors working for teleradiol-ogy companies are regulated when reporting on patients in the UK. All doctors practising in the UK includ-ing those working for teleradiology companies must be registered and hold a current licence to practise with the General Medical Council (GMC)

and are subject to revalidation. How-ever, increasingly doctors based out-side the UK are being employed to report on images of UK patients. The Royal College of Radiologists is con-cerned that these radiologists are not required to be registered with the GMC, have a licence to practise or be subject to a process of revalida-tion. Dr Giles Maskell, RCR Presi-dent said: “The reporting of patients’ images by a local radiologist who can speak directly to other clinicians is the best for patients. Remote report-ing or teleradiology can provide a useful alternative when local services are over-stretched. However, patients deserve the reassurance that quality and safety are always the first pri-ority, whoever reports their images. All radiologists who issue reports on UK patients should be subject to revalidation or an equivalent process regardless of where they are based.”

The RCR statement lists 13 Key Principles for the utilisation and prac-tice of safe, high quality teleradiology including:

• A teleradiology service should always have the safety and well-being of the patient as its first priority. Sec-ondary incentives, financial or other-wise, must always be subsidiary.

• Teleradiology is a medical act and should therefore be governed by the same systems that safeguard patients in all other medical acts.

• Teleradiology must form part of an integrated radiology service and be subject to the same governance frame-work as the rest of the service with all participating radiologists working within a clearly documented quality assurance framework in line with RCR guidance.

• Patients should be clearly informed if their imaging tests are to be reported by a radiologist working outside the service where the images were acquired.

While The Royal College of Radi-ologists recognizes there is a role for teleradiology, it believes this should not replace or destabilise the tradi-tional model of an on-site, local radi-ology service.ROYAl cOllEGE FOR RADIOlOGY lOnDOn. UKwww.rcr.ac.uk

opening of Museum for Medical technology

The Siemens Museum for Medi-cal Technology was recently opened in Erlangen, Germany. Occupying 400 square meters in all, the Med-Museum offers an overview of the development of medical technol-ogy, a field in which Siemens has played a key role for more than 160 years – from X-ray technology to laboratory diagnostics. Important innovations and their inventors are taken as examples, bringing home the history of medical technology to visitors in multimedia format from the field’s inception, in the mid-19th century, to the present day. The historic space once occupied by a machine shop dating to 1893 show-cases selected pieces such as the first X-ray, computed tomography (CT), and magnetic resonance imaging (MRI) systems from Siemens while also providing background informa-tion and explaining how these tech-nologies work. The MedMuseum also traces the development of the various companies that were prede-cessors of Siemens Healthcare. For Joachim Herrmann, Bavarian Minis-ter of the Interior and Building, who opened the new museum, this spe-cific exhibition concept underscores the exceptional position of Siemens in the field of medical technology: “We can be proud that high-tech medical equipment ‘made in Erlan-gen’ has such an excellent reputation all over the world. The new Siemens MedMuseum lets the visitor relive this Erlangen-made success story that now spans already more than a hundred years.” SIEMEnS ERlAnGEn, GERMAnYwww.healthcare.siemens.com


New display for clinical image review

Barco has announced the launch of their Eonis 21” display – a new 21-inch clinical review system offering a 40% higher calibrated luminance and almost twice the contrast ratio of pre-vious generations of displays. With a 2MP resolution for view-ing consistent, quality-controlled images for accurate review, the new display is particularly suit-able for image review in radiology and cardiology and viewing stan-dard PACS images, RIS data, and image-enabled EMR.

The full-screen format – with the traditional 4:3 aspect ratio and a resolution of 1600 x 1200 pixels – has been specifically designed to meet today’s requirements for image review in radiology and cardiology.

With LED backlights, the new display raises clinical performance to a higher level by delivering a DICOM calibrated brightness of 250 cd/m² as well as an excellent contrast ratio (1500:1) for view-ing high-detail images quickly and efficiently and with fewer image manipulations. Built-in ambient light presets ensure perfect image quality in dark or bright view-ing environments. Additionally, thanks to the front-of-screen con-sistency sensor – which automati-cally aligns image quality every time the display is switched on –image consistency is guaranteed over timeBARcOKORTRIJK, BElGIUMwww.barco.com

upDATetecHNoLoGY

Contrast-enhanced mammogra-phy depicts areas in the breast asso-ciated with hypervascularized breast lesions, after an iodinated contrast agent has been injected.

Conventional mammography energy levels are only slightly sensitive to the presence of iodine in the breast. Atypical clinical concentration of iodine in the breast results in a low signal inten-sity, and is hardly distinguishable from the background breast morphology. To obtain images that efficiently highlight iodine, cancellation of the background breast tissue is necessary .

One option to cancelling the back-ground breast tissue and spotlight iodine enhanced areas is Contrast Enhanced Spectral Mammogra-phy (CESM), a technique based on dual-energy acquisitions, where two images are acquired using distinct low-energy (LE – standard mammog-raphy KV and filtration) and high-energy (HE – higher KV with strong filtration) X-ray spectra. The differ-ences between X-ray attenuation of iodine and breast tissues at these two energy levels are exploited to suppress the background breast tissue.

The GE SenoBright Contrast- Enhanced Spectral Mammography (CESM) system allows CESM to be performed as an adjunct to inconclu-sive mammography and ultrasound, highlights areas of unusual blood flow patterns which may be cause for increased suspicion. Using an iodine contrast agent, SenoBright takes two images per view at different X-ray

exposures. It then recombines them to highlight the contrast-enhanced areas— all in a simple, quick, pro-cedure that takes less than ten min-utes— much like a regular mammog-raphy exam. SenoBright is available with the new GE Senographe Essen-tial and Senographe DS systems and as an upgrade for these systems.

The Senobright system features a powerful X-ray tube and generator and a fast-reading digital detector, as well as a proprietary recombination algorithm. Image acquisition is fully automated. The spectral data neces-sary to create two images per view, are acquired automatically. These are a standard mammographic image showing tissue density, and a con-trast-enhanced image in exactly the same position with the background signal subtracted out. GE hEAlThcARE lITTlE chAlFOnT, BUcKS, UKwww3.gehealthcare.com/

Contrast Enhanced Spectral Mammography(CESM) System

The SenoBright functionality is available in a field upgrade on GE’s Senographe Essential and Senographe DS systems.

Illustration of CESM principle on a breast tissue equivalent structured phantom containing disks with typical clinical iodine concentrations. Left Panel. The low energy image is acquired using standard mam-mography energy levels. Note that the iodine is difficult to detect. Mid Panel. The high energy image is acquired using a mean energy above the K edge of iodine. Iodine detectability is improved but still limited by the structured tissue equivalent background. Breast tissue contrast is also deteriorated at these high energy levels. Right panel. In the iodine image, the background tissue is suppressed and the iodine is easily visualized


Mobile c-Arm

The Cios Alpha system from Sie-mens covers the needs of all relevant clinical disciplines, opening up new possibilities in surgical imaging. The system is the first mobile C-arm to use Full View FD technology to provide outstanding image quality and up to 25% more coverage even during image rotation. The system also features the Retina Imaging Chain with IDEAL dose reduction to provide high-quality images at very low dose. Full vascular support is built in featuring unique live graphical overlay

One of the most powerful mobile C-arms available (25 kW), the sys-tem is nevertheless easy to use, with full table-side control and single-touch positioning for effortless operability. SIEMEnS ERlAnGEn, GERMAnY www.siemens .com

3d Image guidance system The recently launched EmboGu-

ide from Philips is the company’s latest innovation in interventional oncology and is designed to treat difficult-to-reach tumors or tumors in patients who are deemed unsuit-able for surgery. Designed for use with Philips’ interventional X-ray

system to perform tumor emboli-zation procedures, EmboGuide is a live 3D image guidance tool that supports the increasing number of minimally-invasive procedures. Such procedures involve the insertion of a catheter, which must be guided to the tumor site with the aid of live image-

guidance. One example of tumor embolization procedure is transar-terial chemo-embolization (TACE), used for palliative treatment of liver tumors. It involves simultaneous local administration of chemother-apy and beads that block the arteries feeding the liver tumor.

Agfa HealthCare has launched its new web-enabled mobile image management technology that brings the power of the company’s ICIS system to iPhones, iPads, Web, and Android mobile digital devices. The new ICIS Mobile and Web Capture enables physicians, caregivers, and patients to publish clinically rel-evant medical images from mobile and web based devices to a securely indexed entry within a patient’s electronic health record (EHR). Doing so, the technology fosters increased physician-to-physician engagement across departments and beyond the hospital’s walls, while creating an environment that empowers patients to become more informed and involved in the care management process.

“Mobile technology is now wide-spread within the healthcare space with physicians frequently using their mobile phones or tablets to record patient data and connect with the EHR,” stated an Agfa spokesman . “ICIS’s new mobile and web capture technology is designed to leverage this trend to improve the delivery of care and reduce cost by uniting the conve-nience and immediacy of mobile computing with the power of the ICIS enterprise imaging platform. Now, ICIS enables clinicians, as well as their patients, to use mobile devices to capture, access, and index medical images and videos within the EHR.”

“ICIS’s easy to use mobile interface also empowers patients to become involved in their medi-cal care – allowing them to take

‘medical selfies’ as needed, reduc-ing unnecessary and costly visits. For instance, with ICIS, a patient being treated for a diabetic foot ulcer now has the ability to pro-duce and share relevant wound images within a highly secure application, enabling the physician to monitor the treatment progress without the need for repeated visits to an examination room.”

ICIS now combines an HTML5 Web-based mobile interface with a flexible workflow and robust meta-data to allow seamless accessibility while providing scalability to meet the needs of various departments within the hospital. Leveraging the power of ICIS, the company’s “gold standard” enterprise imaging platform, and the flexibility of ICIS View, their web-based universal viewer, the new ICIS’s mobile and web capture technology delivers enhanced viewing, sharing, and integration across mobile plat-forms, of images from all sources securely accessible on a single mobile-based viewer.AGFA hEAlThcAREMORTSEl, BElGIUM www.agfahealthcare.com

Mobile and web version of enterprise imaging platform


TeChNOLOGY upDATe

The system has been shown to detect more than twice as many tumor feeding arteries compared to conventional imaging methods such as Digital Subtraction Angiog-raphy (DSA). This allows interven-tional radiologists to optimize the catheter locations for embolization and plan a route to them. During the administration of the emboliza-tion agent, EmboGuide accurately superimposes the planning informa-tion onto the interventional X-ray system’s live images to monitor the treatment progress and determine its endpoint. PhIlIPS EInDhOVEn, nEThERlAnDSwww.healthcare.philips.com

Ultrasound training simulator

MedaPhor, a global provider of advanced ultrasound education and training for medical professionals, has announced that a new study into the impact of ultrasound skills training using its ScanTrainer TVS (Transvaginal) simulator has con-cluded that novices’ performances significantly improved with practice. Their learning curves reached the level of expert performance after between three and four hours of simulator training. The study gives powerful support to the effective-ness of ScanTrainer simulation train-ing in early phase ultrasound skills acquisition

The study was carried out by one of Europe’s leading hospitals, Copenhagen University Hospital Rigshospitalet A group of 16 ultra-sound novices and 12 experienced obstetrics and gynecology consul-tants participated. The first two per-formances of each group, on seven

selected modules on the ScanTrainer TVS simulator, were used to identify valid and reliable metrics. Perfor-mance standards were analyzed and novices were then instructed to con-tinue practicing until they attained the scanning performance level of the expert subgroup.

In the selected female pelvis scan-ning skills, all novices reached the expert ultrasound scanning perfor-mance level after five sessions on the ScanTrainer TVS simulator, an average of 3 hours and 39 minutes of simula-tor-based training.

The study is powerful evidence that training using the ScanTrainer simula-tor can rapidly accelerate ultrasound scanning skills acquisition in a sim-ulated setting in the early learning phase.MEDAPhOR cARDIFF, UK www.medaphor.com/ Line of contrast injectors

VIVID IMAGING have announced the global launch of their Zenith value range contrast product line. Zenith offers injector solutions for MR, CT and Interventional studies. The contrast injectors offer a com-prehensive features with low cost of operation. Zenith injectors offer customized branding options for RX Contrast Manufactures, OEMS and Channel Partners. The Zenith

injector line is currently approved for use in Asia, Europe, Middle East and South America. The company is offering exclusive and non exclu-sive distribution options to regional dealers, OEMS and RX Contrast Manufacturers.VIVID IMAGInGKOWlOOn, hOnG KOnGhttp://technology.vividimaging.com

U-Arm digital Upgrade

Varian’s i5DR digital upgrade will improve workflow and enhance image quality. The i5DR software is a cost effective solution that is fully integrated into industry stan-dard U-arms. The i5DR user interface provides fully integrated control of the u-arm positioner, col-limator and generator allowing for seamless image acquisition. I5DR is optimized for use with Varian’s 17” x 17” (43cm x 43cm) flat panel digital detector. The i5DR upgrade provides significant savings com-pared to the purchase of an entirely new digital system and extends the life and utility of the existing asset. The i5 DR digital imaging system is a suitable choice for dealers and OEMs alike. Offered as either a complete system or a software-only solution, the i5 DR’s built-in flex-ibility allows for easy installation to any new or existing room as well as easy integration to any OEM prod-uct portfolio. VARIAn MEDIcAl SYSTEMS, Inc. PAlO AlTO, cA, USAwww.varian.com

ESC CONGRESS 365

View thousands of videos, slides, abstracts and reports fromESC Congress in a unique digital library.Online, anytime, for free!

Topic: Cardiac Imaging

www.escardio.org/365

ESC Congress 2014Come and see cuttingedge innovation

Village 1: on Cardiac Imaging& e-Technology: www.escardio.org/ESC2014

Best of ESC Congress 2014

Extend your ESC Congress experienceon Thursday 4 September20:00-21:00 (CET)

Register now for free to watchTHE online event!www.escardio.org/bestofesc2014

European Society of Cardiology

#esccongress escardiodotorg

29 Aug - 2 Sept

WWW.ESCARDIO.ORG/ESC2015

ESC CONGRESS 365

View thousands of videos, slides, abstracts and reports fromESC Congress in a unique digital library.Online, anytime, for free!

Topic: Cardiac Imaging

www.escardio.org/365

ESC Congress 2014Come and see cuttingedge innovation

Village 1: on Cardiac Imaging& e-Technology: www.escardio.org/ESC2014

Best of ESC Congress 2014

Extend your ESC Congress experienceon Thursday 4 September20:00-21:00 (CET)

Register now for free to watchTHE online event!www.escardio.org/bestofesc2014

European Society of Cardiology

#esccongress escardiodotorg

29 Aug - 2 Sept

WWW.ESCARDIO.ORG/ESC2015

Medical reporting

at a whole new level

Learn about Agfa HealthCare at www.agfahealthcare.com

Creating medical reports is a key responsibility throughout the continuum of care, taking up a lot of time from all healthcare professionals. Making transcripts from recordings, including verification and correction, is a slow process that ties up staff resources. By turning speech and relevant data directly into structured input, report turnaround time decreases, the possibility of errors reduces and staff resources are freed up for new ways of working throughout the hospital.

C

M

Y

CM

MY

CY

CMY

K

M0934_Agfa_ad_reporting.pdf 1 30/06/14 15:20

die junejuly 2014

Documents

radiology ultrasound

imaging studies

g n o s i s t e c h

o g y t h e r

e u r o p ejuly

ultrasound images

primordial role of radiology

ultrasound scanning