Basic Principles ofCone Beam Computed

Tomography

Kenneth Abramovitch, DDS, MS*, Dwight D. Rice, DDS

KEYWORDS

� Cone beam computed tomography � Flat-panel silicon detector� DICOM viewer software � Beam-hardening artifacts

KEY POINTS

� The use of cone beam computed tomography (CBCT) imaging in the dental profession hasblossomed since its inception 15 years ago. CBCT unit design has undergone manychanges that enhance CBCT access and practical utility in dentistry. The scanners havebecome smaller, scan patients in an upright position, use primarily flat panel detectors,and readily convert projection data to DICOM file formats. Units themselves have variousscanning options that include the size of the area to be scanned (field of view [FOV]), voxelsize (spatial resolution), bit depth (contrast resolution), and scan times (frame rate).

� CBCT manufacturers have incorporated various aspects of imaging technology in a cost-effective, efficient, and practical manner. There are now numerous CBCT applications inmany software formats that are helpful in a multitude of dental disciplines including but notlimited to dentoalveolar disease and anomalies, vertical root and dentin fractures, jaw tu-mors, prosthodontic evaluations, and advances in orthodontic/orthognathic and implantpatient evaluations. The latter also include mechanisms for surgical and prosthodonticsplint design and the capability of CBCT scan data to bridge with computer-aideddesign/manufacturing image files for the fabrication of various dental restorations.

� Streaking and beam hardening remain as ominous imaging artifact that compromiseCBCT utility in various case situations. However, because of the popularity of CBCT, com-puter hardware and software developers, machine manufacturers and dental researcherswill continue to improve the applications of this imaging modality for the betterment of pa-tient care.

INTRODUCTION

Imaging with cone beam technology has rapidly become a popular and frequentlyused imaging modality to assist dentists and other health care professionals in a multi-tude of diagnostic tasks to improve patient care.Cone beam imaging technology is most commonly referred to as cone beam

computed tomography (CBCT). The terminology “cone beam” refers to the conical

Loma Linda University School of Dentistry, 11092 Anderson Street, Loma Linda, CA 92354, USA* Corresponding author.E-mail address: kabramovitch@llu.edu

Dent Clin N Am 58 (2014) 463–484http://dx.doi.org/10.1016/j.cden.2014.03.002 dental.theclinics.com0011-8532/14/$ – see front matter � 2014 Elsevier Inc. All rights reserved.

Abramovitch & Rice464

shape of the beam that scans the patient in a circular path around the vertical axis ofthe head, in contrast to the fan-shaped beam and more complex scanning movementof multidetector-row computed tomography (MDCT) commonly used in medicalimaging.First introduced at the end of the millennium,1,2 CBCT heralded a new dental tech-

nology for the twenty-first century. Its practical applications for implant dentistry andorthognathic/orthodontic patient care were the main applications at that time. Owingto the dramatic and highly positive impact that CBCT had on these disciplines, addi-tional applications for this technology became apparent. New software programswere developed to improve the applicability and access of CBCT for the care of dentalpatients.Two factors played a big part in the rapid incorporation of CBCT technology into

dentistry, the first of which was the availability of improved, rapid, and cost-effective computer technology. The second was the ability of software engineers todevelop multiple dental imaging applications for CBCT with broad diagnosticcapability.

CBCT VERSUS COMPUTED TOMOGRAPHY

CBCT, by virtue of the terminology, is a form of computed tomography (CT). In a singlerotation, the region of interest (ROI) is scanned by a cone-shaped x-ray beam aroundthe vertical axis of the patient’s head. Digitized information of objects in the ROI suchas shape and density is acquired from multiple angles. These imaging data are thenprocessed by specialty software that ultimately constructs tomographic images ofthe ROI in multiple anatomic planes, namely the standard coronal, axial, and sagittalanatomic planes (Fig. 1) and their various paraplanar derivatives, the parasagittal, par-acoronal and para-axial planes.The historically standard and more sophisticated form of CT, present since the

1970s, was developed in part by British engineer and Nobel Prize winner Sir GodfreyHounsfield. It is of interest that by the end of the decade, the technology of Houns-field’s first scanner was followed by the development of a larger body scanner by agroup of researchers in the United States headed by American dentist and physicistRobert S. Ledley.3 This more advanced form of CT is known as MDCT, although otherterms such as multislice CT and multirow CT are used. Because MDCT is morecommonly used in medicine, it is often referred to as medical CT. However, this

Fig. 1. Standard anatomic planes of imaging used for multiplanar reconstructions in conebeam computed tomography (CBCT) and multidetector-row computed tomography.(Modified from Washington CM, Leaver DT. Principles and practice of radiation therapy.Philadelphia: Mosby; 2004.)

Basic Principles of Cone Beam CT 465

term is a misnomer, as CBCT is now also being used and further modified for patientevaluations in medicine.4,5 A more appropriate term for MDCT might be “conventionalCT.” Differences between CBCT and MDCT have been widely reported.6–9 However,owing to the specific advances and innovations of CBCT technology for the care ofdental patients, it has become and will remain a vital and significant imaging modalityin dentistry.

HISTORICAL DEVELOPMENT OF CBCT UNITS

During the early development of CBCT, the technology was being advanced primarilyfor the dental office. Subsequently, many of the earlier units were modified to includedesigns that more readily fit within dental offices and clinics. The integration of CBCTimaging in dentistry has in some ways paralleled the transition of panoramic imagingx-ray machines into dental offices. Early panoramic units were mainly sit-down,10,11

but there was also a lay-down unit.12 Several other sit-down machines were manufac-tured, but eventually units were made whereby the patient could stand upright for thepanoramic exposure. Upright machines became preferable, as it is more convenientand takes less time to transfer patients into and out of these stand-up panoramic units.The physical size and shape of CBCT units has paralleled this panoramic pathway.

One of the very first commercially available cone beam machines, the NewTom 9000(QR srl, Verona, Italy), was a large unit that scanned the patient lying in a supine posi-tion. It was followed by the NewTom 3G (Fig. 2A). These early NewTom units eventuallylost favor to smaller, sit-down chair units or to stand-up units. These smaller units withbetter scanner quality more readily fit into dental office space and overhead budgets(see Fig. 2B–F). Despite the previous drawbacks of the NewTom prototypes, CBCTunits that scan patients in a supine position have made a comeback; the NewTom5G (QR srl) and the SkyView (MyRay, Imola, Italy) are currently available. These units,with upright patient loading and supine position for patient scanning, are presented inFig. 2G–H. NewTom is also producing standing machines such as the VGi.

EFFECT OF FIELD OF VIEW ON SCANNER TYPE

The size of the scanned object volume is called the field of view, commonly abbrevi-ated as FOV. The FOV for units with a flat-panel detector (see later description) is acylindrical shape in the center of the scanner between the detector and the x-raysource. The CBCT scanning controls are programmed to scan an FOV of sizes andareas that are built into the scanner by the manufacturer. Other factors that affectthe FOV are the size and type of the detector and the degree of beam collimationon the x-ray tubehead. Fig. 3A demonstrates how the dimensions of a flat-paneldetector’s FOV cylinder are expressed by the height of the cylinder (H) and the diam-eter of the base (D). The FOV is a very flexible option in contemporary scanners. Therange of commercially available FOVs for flat-panel detectors can be from 3.0 cm (H)�3.0 cm (D) to 24 cm (H) � 16.5 cm (D) (Table 1).The FOV for image-intensifier detectors is shaped differently, not as a cylinder but

rather as a sphere. The dimensions are usually measured by the diameter of the circu-lar shape in inches (eg, 600, 900, 1200).The size of the FOV significantly affected the evolution of the CBCT scanner. Early

CBCT units were restricted to a single-size FOV that was either large or small, whichlimited the usefulness of the scanner. The general rule was the larger the FOV, thegreater the cost of the scanner. The higher cost is attributed to the larger detectorsize and the larger kilovoltage (kV) generator needed for imaging denser parts of theskull for orthognathic and orthodontic evaluations. The FOV most typically included

Fig. 2. (A) NewTom 3G. This supine CBCT scanner was one of the first commercially availableunits in North America. It was replaced by units that scanned patients seatedwith the head inan upright position. (B) The Accuitomo 170 (J. Morita USA, Irvine, CA). (C) The Scanora 3Dx(Soredex,Milwaukee,WI). (D) The CS 9300 (CarestreamHealth, Rochester, NY). (E) The Ortho-phos XG 3D (Sirona USA, Charlotte, NC). (F) The i-CAT FLX (Imaging Sciences International,Hatfield, PA). (G) The NewTom 5G in patient entry (left) and patient scan (right) positions.This unit is currently manufactured by QR srl, Verona, Italy. (H) The SkyView CBCT scanner(MyRay, Imola, Italy) in patient entry (left) and patient scan (right) chair positions.

Abramovitch & Rice466

the jaws, midface, and skull base. Some had options that included a more extendedpart of the skull toward the vertex, that is, 40 cm (H). Because of the limited indicationsand increased cost, the larger FOVs were not as popular for limited dentoalveolar ap-plications. Smaller FOV units large enough to image 2 to 4 teeth of a jaw (either maxillaor mandible) was another earlier scanner option. The area covered in these smallervolumes is adequate for a thorough 3-dimensional (3D) periapical evaluation ofselected teeth, alveolar bone, and a limited amount of maxillary or mandibular basalbone. Contemporary scanners are now capable of a range of FOVs (see Fig. 3B)from the smaller 3.0 cm (H) � 3.0 cm (D), to the midrange FOVs for coverage of oneor both jaws, to the larger FOVs that include the cervical spine, jaws, more of the para-nasal sinuses, skull base, and parts of the cranium. Larger FOVs that include superiorareas of the skull are not usually indicated for most dental applications.Because of these technological improvements and enhancements, CBCT is now

readily identified as part of the imaging equipment in modern dental clinics. Table 1lists many of the currently available CBCT units with larger FOV capabilities alongwith notations of some of their other options. Table 2 is a similar listing of units with

Fig. 2. (continued)

Basic Principles of Cone Beam CT 467

medium to smaller FOV options. Because of the constant modifications in CBCT scan-ner technology, manufacturers, and machine trade names, the information in thesetables is time sensitive and only current at the time of publication. For additional infor-mation, other listings may also be referenced.13–15

FEATURES OF THE IMAGING PROCESSImage Capture

As in any radiographic imaging system, CBCT requires x-ray production, x-ray atten-uation by an object, signal detection, image processing, and image display. These

Fig. 3. (A) Cylindrical shape and measurement characteristics of the field of view (FOV)for CBCT. (B) The different FOV option sizes from the Vatech CBCT (Vatech America,Fort Lee, NJ). Many CBCT units now have the capability of scanning a range of FOV sizes.

Abramovitch & Rice468

parameters are vital to all aspects of dental imaging, but they are understandably moresophisticated for CBCT.During a rotational scan of an object, multiple exposures are taken at fixed intervals

(angles) of the rotation. Each of these exposures is referred to as a basis image. Theimages are standard radiographic images captured on the detector, and the signal ofeach projection is unique for each of the different angles in the rotational arc. Instan-taneously the image data for each basis image are sent to a data-storage area so thatthe detector can be cleared to capture the next basis image at a position intervalfurther along the rotational arc. Once the rotation is complete and all the basis im-ages are made, the complete set of basis images forms the “projection data.” Thetotal number of basis images taken depends on the radiographer’s preferencesand the scanner’s capability. This total ranges from 100 to 600 basis images perscan. The greater the number of basis images, the longer the scan time, the greaterthe radiation dose, and the better the quality of the constructed images. Fig. 4 dem-onstrates a hypothetical scheme for projection-data formation with the capture of 2basis images.

Imaging Software and Data File Management

Image reconstruction software programs, usually proprietary to each machine manu-facturer, then manage the projection data and construct a 3D volumetric data set.These processed data are then accessed to construct various types of images fordisplay. The choice of images constructed depends on the power of the imaging soft-ware and the needs and preferences of the clinician. The image selection from 3D soft-ware is not limited to a single type of image display. Depending on the capability of thesoftware, there are multiple options of image construction from the 3D volumetric dataset. Most scanner programs display a primary image reconstruction of the object inthe 3 anatomic planes of imaging: the axial, sagittal, and coronal planes. These pri-mary reconstruction displays are also referred to more typically as the multiplane ormultiplanar images. Primary multiplanar reconstructions from 2 different software pro-grams are presented in Fig. 5. The same volumetric data set can be used to alsoconstruct multiple kinds of secondary reconstructions. The choice of secondary

Basic Principles of Cone Beam CT 469

reconstruction is often task specific, and is also related to the reconstruction optionswithin the scanner’s proprietary software.At present, a variety of independent third-party imaging software is commercially

available for image reconstruction of CBCT volumetric data sets. Third-party imagingsoftware is software not associated with the capture and proprietary software of theCBCT scanner. A limited selection of third-party software is listed in Table 3.If third-party software is being used, the file format of the volume set must be con-

verted from the proprietary file format or file language to a more universal or commondigital file format. This common format must be conformant with the Digital Imagingand Communications in Medicine standard (DICOM 09v11dif); that is, the currentDICOM standardized file format.16 This digital format is the International Organizationfor Standardization (ISO) referenced standardized digital file format for medical im-ages and related information, namely ISO 12052. To facilitate access to health care,multiple imaging modalities (x-ray, visible light, ultrasound, and so forth) used in med-icine and dentistry must be compliant with ISO 12052. Digital applications in veterinarymedicine also follow this standard.If CBCT vendors do not specifically use the DICOM file format in their proprietary

scanner software, their proprietary software should have the capability of convertingthe volume data to the DICOM standard file format. In so doing, they make their vol-ume data usable in other and often more specialized software applications.Types of task-specific reconstruction capabilities of viewing software include, but

are not limited to, panoramic reconstructions, implant planning reconstructions with2-dimensional and 3D windows, temporomandibular joint reconstructions, airway re-constructions, and so forth. Examples of these latter reconstructions are presented inFig. 6.

X-Ray Tube and Generator Systems

Because CBCT is a radiographic imaging system, scanners have x-ray tubes with kVand milliamperage exposure controls. Although the time of exposure is usually anexposure control for an x-ray system, in CBCT the time of the exposure is actuallydependent on the number of basis images and the degree of spatial resolutionrequested in the voxel size. The smaller the voxel size and the greater the numberof basis images, the longer the exposure. The major difference in a CBCT exposurecompared with the exposure of intraoral and panoramic imaging is that the CBCTexposure consists of capturing the series of multiple basis images. Because of theprocess of basis-image projection, the x-rays are not generated during the entire rota-tional path. In most units, the exposure is pulsed at intervals so that there is time be-tween basis-image acquisition for the signal to be transmitted from the detector areato the data-storage area and the detector to rotate to the next site or angle of expo-sure. Hence, the x-ray tube does not generate x rays for the entire rotational cycle.These intervals may inherently reduce patient exposure during the time interval thatthe detector is not ready to receive x rays. These intervals are also beneficial for thex-ray duty cycle, reducing heat buildup during an exposure cycle.In general, the longer the exposure and the more basis images produced, the longer

it takes to complete the rotational arc. This time for the acquisition of basis images isknown as the frame rate. For a shorter exposure, the rotational arc remains the samebut the frame rate is reduced. In this scenario where less basis images are taken, theradiation exposure is less, the rotational arc takes less time, and the scanner partsrotate faster. The clinician can actually observe the slower or longer scan times neces-sary for longer exposures with higher frame rates.

Table 1Scanners with large field of view (FOV)

Model Manufacturer

Maximum toMinimum FOVHeight 3 Width (cm)

MinimumVoxel (mm3)

Two-DimensionalOptions Bit Depth

ScanTime (s) kV Notes

3D Accuitomo 170 J. Morita USA, Irvine,CA, USA

12 � 17 to 4 � 4 0.08 No 14 5.4–30 90 —

3D eXam KaVo, Charlotte, NC,USA

17 � 23 to 8 � 8 0.125 a 14 8.5–26 120 —

CS 9300 Carestream Health,Rochester, NY, USA

13.5 � 17 to 5 � 5 0.09 Panoramiccephalogram

14 12–20 90 Previously Kodak(dental division)

Available in 2 versions

DaVinci D3D Cefla Dental, Imola,Italy

15 � 15 to 7 � 7 0.17 a 12 a a Supine position for scan

Galileos ComfortPlus

Sirona USA, Charlotte,NC, USA

Sphere 15.4 cm (D) 0.125 No 12 14 98 Plus model is upgradedComfort

Uses image intensifierdetector

i-CAT FLX Imaging SciencesInternational,Hatfield, PA, USA

17 � 23 to 8 � 8 0.125 Panoramic 14 5–26.9 120 QuickScan1 optionallows for ultralowdose exposures

NewTom 5G QR srl, Verona Italy 16 � 18 to 6 � 6 0.075 No 14 18–26 110 Supine position for scan

NewTom VGi QR srl, Verona, Italy 15 � 15 to 6 � 6 0.075 No 14 18–26 110 VGi Flex versionintended for mobileuse

Abramovitch

&Rice

470

PaX-Reve3D Plus Vatech America Inc,Fort Lee, NJ, USA

15 � 19 to 5 � 5 0.08 Panoramic 14 15–24 90 —

ProMax 3D Max Planmeca USA Inc,Roselle, IL, USA

17 � 22 to 5.5 � 5 0.10 Panoramic 15 18–26 96 Can obtain a stitched 26� 23 cm FOV andupgradable toProFace 3D Photos

ProMax 3D Mid Planmeca USA Inc,Roselle, IL, USA

17 � 20 to 5 � 4 0.10 Panoramiccephalogram

15 18–26 90 —

Quolis Alphard3030

Asahi Roentgen Ind.Co., Ltd, Kyoto, Japan

17.9 � 20 to 5.1 � 5.1 0.10 No a 17 110 —

Scanora 3D Soredex, Milwaukee,WI, USA

13 � 14.5 to 6 � 6 0.133 Panoramic 12 10–26 90 —

Scanora 3Dx Soredex, Milwaukee,WI, USA

24 � 16.5 to 5 � 5 0.10 Panoramic a 18–34 90 —

SkyView MyRay, Imola, Italy Sphere 22.9 cm (D)15.3 cm (D)10.2 cm (D)

0.17 No 12 10–30 90 Supine position for scanUses image-intensifierdetector

WhiteFox Acteon North America,Mt. Laurel, NJ, USA

17 � 20 to 6 � 6 0.10 No 16 18–27 105 —

a Information was not available at press time.

Basic

Prin

ciplesofConeBeam

CT

471

Table 2Scanners with medium and small FOV

Model Manufacturer

Maximum toMinimum FOVHeight 3 Width (cm)

MinimumVoxel (mm3)

Two-DimensionalOptions

BitDepth

ScanTime (s) kV Notes

AUGE ZIO Asahi Roentgen Ind.Co., Ltd, Kyoto,Japan

8 � 10 to 5.5 � 5 0.1 Panoramiccephalogram

12 8.5–17 95 —

Cranex 3D Soredex, Milwaukee,WI, USA

6 � 8 to 6 � 4 0.085 Panoramiccephalogram

a 11–17 85 —

CS 9000 3D Carestream Health,Rochester, NY, USA

3.75 � 5Stitched7.5 � 3.75

0.076Stitched 0.2

Panoramiccephalogram

14 10.8 90 Will stitch 3 scanstogether

Finecube XP62 Yoshida, Tokyo, Japan 7.5 � 8.1 0.1 No 14 8.6–34 90 Marketed as PreXion3D by PreXion Inc inthe USA

GXCB-500 Gendex, Hatfield, PA,USA

8 � 14 to 8 � 8 0.125 No 14 8.9–23 120 Powered by i-CAT

GXCB-500 HD Gendex, Hatfield, PA,USA

8 � 14 to 2 � 8 0.125 Panoramic 14 a 120 Powered by i-CAT

GXDP-700 Gendex, Hatfield, PA,USA

6 � 8 to 6 � 4 0.2 Panoramiccephalogram

a 10–20 90 —

i-CAT Precise Imaging SciencesInternational,Hatfield, PA, USA

8 � 14 to 2 � 8 0.125 Panoramic 14 4–23 120 —

I-Max Touch 3D Owandy, Croissy-Beaubourg, France

8.3 � 9.3 0.156 Panoramiccephalogram

8–16 20 86 —

Abramovitch

&Rice

472

OrthopantomographOP300

Instrumentarium,Milwaukee, WI,USA

6 � 8 to 6 � 4 0.085 Panoramiccephalogram

14 10–20 90 —

Orthophos XG-3D Sirona USA, Charlotte,NC, USA

8 � 8 to 5 � 5 0.1 Panoramiccephalogram

12 14 90 —

PaX-i3D Green Vatech America, FortLee, NJ, USA

10 � 16 to 5 � 5 0.12 Panoramiccephalogram

14 5.9a 100 —

PreXion 3D Eclipse PreXion, Inc, SanMateo, CA, USA

8 � 11 to 8 � 7.5 0.15 Panoramiccephalogram

14 8.7–17.4

90 —

PreXion3D Elite PreXion, Inc., SanMateo, CA, USA

8 � 7.5 to 5.6 � 5.2 0.11 Panoramic 13 8.6–33.5

90 —

Promax 3D Classic Planmeca USA Inc,Roselle, IL, USA

8 � 8 to 8 � 4 0.1 Panoramiccephalogram

15 18 90 —

Promax 3D Plus Planmeca USA Inc,Roselle, IL, USA

9 � 14 to 5 � 4 0.1 Panoramiccephalogram

15 18 90 —

Promax 3D s Planmeca USA Inc,Roselle, IL, USA

8 � 5 to 5 � 5 0.1 Panoramiccephalogram

15 18 90 —

Suni3D Suni Medical Imaging,San Jose, CA, USA

5 � 5 to 5 � 8 0.08 Panoramiccephalogram

16 15–24 90 —

Veraviewepocs J. Morita USA, Irvine,CA, USA

8 � 10 to 4 � 4 0.125 Panoramiccephalogram

13 9.4 90 Variousconfigurationsavailable

a Information was not available at the time of writing.

Basic

Prin

ciplesofConeBeam

CT

473

Fig. 4. Basis-image capture for a hypothetical CBCT rotational scan of the cervical spine. Twobasis-image capture sequences are depicted in this diagram as the machine rotates counter-clockwise from Position 1 to Position 2. An arrow depicts the counter-clockwise rotation.CBCT scans routinely capture in the range of 100 to 600 basis images per rotational scan.(Modified from Zhen X, Yan H, Zhou L, et al. Deformable image registration of CT and trun-cated cone beam CT for adaptive radiation therapy. Phys Med Biol 2013;58(22):7979–93.)

Abramovitch & Rice474

A data set with fewer basis images may be undersampled. Undersampled data setshave a lower signal-to-noise ratio, and thus lack the range of image contrast capablewith a more complete volumetric data set. However, depending on the diagnostictask, the degree of image degradation from these smaller data sets with shorter expo-sures may still be adequate for certain diagnostic tasks. This shorter feature reducespatient exposure and is particularly helpful in reducing motion artifact in scans ofyounger patients, geriatric patients, or those with disabilities, during which motion arti-fact is more difficult for the patient to control. These smaller data sets also have lesscomputational and construction time. With fewer data, they require less storagespace. An example of undersampling is illustrated in Fig. 7.Exposure factors for a CBCT scan can be preset from an exposure selection guide,

or can be determined by automated features in the image-acquisition software from ascout exposure. Some units may have a direct automated exposure feedback featurein the detector that determines the exposure factor for more optimal signal detection.The automated exposure control predates CBCT technology and has been availablesince the introduction of charge-coupled device sensors for digital panoramic radiog-raphy. CBCT units that scan larger object areas (larger FOV) generally need higher kVpotentials. The higher kV is necessary for adequate penetration of denser and largeranatomic structures in the maxilla, midface, and skull base. Consequently, higherkV is often necessary for adequate diagnostic quality of the larger FOV data sets.

Fig. 5. Examples of multiplanar reconstructions. The upper example (A) is constructed byOne Volume viewer software (J. Morita USA). The lower (B) reconstruction is by CS 3D Im-aging Software (Carestream Health, New York).

Basic Principles of Cone Beam CT 475

Table 3Third-party software available for imaging CBCT data sets

Software Manufacturer Uses

CS 3D Carestream Dental, Rochester, NY, USA Multiple applications

Dolphin 3D Dolphin Imaging and ManagementSolutions, Chatsworth, CA, USA

Multiple applications

EasyGuide Keystone Dental, Burlington, MA, USA Implant planning

InVivoDental Anatomage Inc, San Jose, CA, USA Multiple applications

OnDemand3D Cybermed Inc, Irvine, CA, USA Multiple applications

OsiriX Pixmeo SARL, Bernex, Switzerland Multiple applications

Procera Software Nobel Biocare USA, LL, Yorba Linda, CA, USA Implant planning

Ultra-Fast CBCTReconstructionSoftware

Bronnikov Algorithms, The Netherlands Multiple applications

Fig. 6. Examples of secondary reconstructions from various CBCT software programs. (A)Two-dimensional (2D) panoramic reconstruction. Although a CBCT scan is not indicatedsolely for panoramic imaging, many imaging software packages can reconstruct panoramicimages from the storage data. (B) Implant planning with 2D reconstructions and a tracing ofthe mandibular nerve. (C) Implant planning with 2D/3-dimensional (3D) reconstructions. (D)Bilateral reconstructions of the temporomandibular joints in coronal and sagittal sections.(E) Sagittal reconstruction without (top) and with (bottom) Airway Measurement toolfrom InVivo 5.2 imaging software (Anatomage, San Jose, CA). When the airway is tracedin the airway measurement window, the program wizard computes the volume of theairway space. Threshold values for compromised airway volumes have not yet been deter-mined for this software.

Abramovitch & Rice476

Fig. 6. (continued)

Basic Principles of Cone Beam CT 477

Image Sensor Systems

Two types of image detectors are used as the sensors in contemporary CBCT units. Ascanner will have either (1) a charge-coupled device with a fiber-optic image intensi-fier, or (2) an amorphous silicon flat-panel detector. Examples are presented in Fig. 8.During the initial introduction of CBCT, most units were constructed with the large,bulky image-intensifier detectors. In the latter half of the first decade of commercialCBCT development, CBCT scanners have nearly all transitioned to the smaller, flat-panel linear array detectors. However, as noted in Tables 1 and 2 that list represen-tative CBCT imaging systems, Sirona and MyRay still manufacture scanner unitswith this type of detector.

Fig. 6. (continued)

Fig. 7. (A) Sagittal temporomandibular joint reconstruction from projection data processedfrom a full quota of basis images in the projection data set. (B) Sagittal temporomandibularjoint reconstruction from a shorter exposure scan that has fewer basis images in the projec-tion data and resulting volumetric data set. There is less detail and contrast resolution in theresulting image display than with projection data from a full quota of basis images used forconstruction of the volumetric data set.

478

Fig. 8. Two CBCT scanners from Sirona USA. The Galileos (top) has a charge-coupled image-intensifier detector. The Orthophos XG 3D unit (bottom) has the smaller flat-panel detector.The detectors are demarcated with dotted outlines. Differences between the two aredescribed in the text.

Basic Principles of Cone Beam CT 479

The image-intensifier detectors are larger and make the scanners’ overall dimen-sions larger, which may be critical for certain office designs. In addition to beingmore sensitive and susceptible to distortion from magnetic fields, image displaysfrom these detectors also demonstrate greater distortion of the grid dimensionswhen moving away from the center of the detector (Fig. 9A), which ultimately reducesmeasurement accuracy of the reconstructed images.17 Because of their sensitivity tomagnetic fields, the image-intensifier detectors require more frequent calibration. Inaddition, the phosphors in image intensifiers lose their sensitivity over time and use,and the entire image-intensification unit may need to be replaced to maintain imagequality. This process is very expensive.18 Despite the drawbacks, in certain casesthe data sets from these detectors are more compatible with “bridging” to some ofthe data sets used in computer-aided design and manufacturing (CAD/CAM) technol-ogy, and thus remain useful.The flat-panel detectors are thin, amorphous silicon transistor panels with a cesium

iodide scintillator. The scintillator is the part of the detector used to amplify the elec-trical signal from the x-ray attenuation. Besides being smaller and less bulky, the flat

Fig. 9. Distortion patterns produced by image detectors. (A) Grid type X is the type of grid-distortion pattern produced by the image-intensifier detector that affects the image con-struction and is subsequently noted in the image display. There is distortion of the imagegrid when moving away from the center. (B) With flat-panel detectors (ie, grid type Y)the image receptor area receiving the signal from the flat-panel detector’s scintillator isflat. Therefore, even at more distant areas from the center of the grid, there is minimalto no distortion of the grid pattern.

Abramovitch & Rice480

panels have minimal distortion of the image dimensions at the periphery of an imagedisplay (see Fig. 9B); hence, these units are considered to generate better data sets.Because these detectors are smaller than their image-intensifier predecessors, CBCTunits with the flat-panel detectors have smaller footprints. This feature alone hadmadethe flat-panel detector more popular. Differences in image quality between these de-tectors are shown in Fig. 10.Another property of the image detector is the bit depth, an exponential binary prop-

erty expressing the total number of gray shades the detector is able to discriminate. A14-bit detector (ie, 214) can display 16,384 shades of gray. The range of bit depth ofcommercial CBCT units ranges between 12 and 16 bits (see Tables 1 and 2), indi-cating the wide range of contrast discrimination capability. Although the detector iscapable of this degree of gray-scale discrimination, limiting features to the contrastresolution include the lower bit depth of the imaging software and the monitor display,

Fig. 10. Comparative reconstructions of two different scans of the same posterior left maxil-lary quadrant from a scanner with a flat-panel detector (left) and one from a charge-coupled device image intensifier (right). The improved image quality and the highersignal-to-noise ratio are noted in the left image. (Courtesy of Dr Bruno Azevedo, WesternUniversity, Pomona, CA.)

Basic Principles of Cone Beam CT 481

and the visual perception of the viewing clinician. Even though bit depth is importantfor contrast resolution, the American College of Radiology has concluded that there isno added benefit to diagnostic interpretations by the use of higher than 8-bit depth inthe workstation’s operating system.19

Scatter and Beam-Hardening Artifact

Scatter and beam-hardening artifact occurs in CT imaging where image reconstruc-tions of a data set are necessary for review of the data volume. Dense metal structuresfrequently in the FOV for dental applications present metal artifact on CBCT recon-structions. Silver amalgam, precious and semiprecious metal alloys used in coronalrestorations, dental implants, silver-point endodontic fillings, and, to a lesser extent,gutta percha endodontic fillings, all create these artifacts in image reconstructions.The artifact presents as light or dark streaks, or as a dark periphery adjacent tometallic borders. Scatter artifact is seen as radiopaque lines and patterns of metallicdensity that “scatter” on image reconstructions. The main types of beam hardeningare the dark streaks or dark bands that show up in the image reconstructions. Thelatter often simulate disease such as recurrent caries or fractures in endodonticallytreated teeth. The light streaks often superimpose regular anatomy, and may alsosignificantly degrade image quality.These artifacts are prominent problems for dental applications with CBCT, as

metallic restorations are often within the FOV of most CBCT scans of dental patients.The metallic restorations then cause the resultant beam hardening and streak artifact,which then compromises the image quality with the various areas of dark and light arti-fact. Fig. 11 illustrates examples of how these artifacts degrade image quality andmake image assessments difficult.Recent attempts via software correction algorithms have been reported that have

the potential to control these visible artifacts on image reconstruction.20,21 However,the application of software correction modes to reduce these artifacts have been infe-rior to noncorrected software viewing programs when evaluating peri-implant and

Fig. 11. Beam hardening and streak artifact in CBCT image reconstructions. (A) Axial sectionwith dental implant in #18 region highlighted by black arrow. (B) Beam-hardening artifact isindicated by red arrows. The green arrows depict streak artifact. (C) The locations of cross-sectional and parasagittal reconstructions are shown. (D) The effect of beam hardeningsimulating peri-implantitis and alveolar bone defects in the cross-sectional and parasagittalreconstructions. (E) The effect of streak artifact creating the outline of a “ghost” implant(as well as other radiopaque streak outlines) in the cross-sectional reconstruction. The streakartifact makes it more difficult to discern the validity of the cortical bone outlines. (Courtesyof Dr. Gerald Marlin, Washington, DC.)

482

Basic Principles of Cone Beam CT 483

periodontal disease22 as well as root fractures.23 Consequently, there are no immedi-ate methods to correct or minimize these prominent artifacts. The best way to avoidstreaking and beam hardening is to try to keep the FOV as small as possible in anattempt to minimize or keep these metals outside the FOV. In so doing, one may beable to minimize their impact on image reconstructions.

SUMMARY

CBCT is now a well-accepted diagnostic tool for the care of dental patients. Designchanges in the evolution of contemporary CBCT scanners include making the unitssmaller, and making changes whereby instead of needing to be scanned in a supineposition, the patient either sits or stands upright during the scan. Along with thesedesign changes, better stabilization devices for the patient’s head and chin were pro-duced. Mechanical changes included the switch to smaller, flat-panel silicon detectorswith better image quality compared with the bulkier, cumbersome, and eventuallymore costly image-intensifier detectors.Variable kV and multiple options for voxel size, FOV dimensions, scan times, and so

forth, then followed, alongside more powerful software applications for the care ofdental patients. The ability of CBCT manufacturers to use various aspects of imagingtechnology in a cost-effective, efficient, and practical manner means that there arenow numerous CBCT applications that are helpful in a multitude of dental disciplines.These applications include, but are not limited to, dentoalveolar abnormality, verticalroot fractures, jaw tumors, prosthodontic evaluations, and advances in orthodontic/orthognathic and implant patient evaluations. The latter also include mechanismsfor surgical and prosthodontic splint design and the capability of CBCT scan datato bridge with CAD/CAM image files for fabrication of various dental restorations.This approach facilitates implant and prosthodontic rehabilitation by synchronouslyplanning and subsequently milling coronal restorations for teeth and rootform im-plants. As the demand for CBCT technology continues to increase, so will the numberof new applications for improved diagnostic techniques.

REFERENCES

1. Mozzo P, Procacci C, Tacconi A, et al. A new volumetric CT machine for dentalimaging based on the cone-beam technique: preliminary results. Eur Radiol1998;8(9):1558–64.

2. Arai Y, Tammisalo E, Arai Y, et al. Development of a compact computed tomo-graphic apparatus for dental use. Dentomaxillofac Radiol 1999;28(4):245–8.

3. Sittig DF, Ash JS, Ledley RS. The story behind the development of the first whole-body computerized tomography scanner as told by Robert S Ledley. J Am MedInform Assoc 2006;13(5):465–9.

4. Miracle AC, Mukherji SK. Conebeam CT of the head and neck, part 1: physicalprinciples. AJNR Am J Neuroradiol 2009;30(6):1088–95. http://dx.doi.org/10.3174/ajnr.A1653.

5. Miracle AC, Mukherji SK. Conebeam CTof the head and neck, part 2: clinical ap-plications. AJNR Am J Neuroradiol 2009;30(7):1285–92. http://dx.doi.org/10.3174/ajnr.A1654.

6. Miles DA. Color atlas of cone beam volumetric imaging for dental applications.Chicago: Quintessence Publishing; 2007. p. 9–14.

7. MacDonald-Jankowski DS, Orpe EC. Computed tomography for oral and maxil-lofacial surgeons. Part 2: cone-beam computed tomography. Asian J Oral Max-illofac Surg 2006;18:85–92.

Abramovitch & Rice484

8. Scarfe WC, Farman AG. Cone-beam computed tomography. In: White SC,Pharoah MJ, editors. Oral radiology principles and interpretation. St Louis(MO): Mosby-Elsevier; 2009. p. 225–43.

9. Drage N, Rout J. Diagnostic imaging. In: Kau CH, Richmond S, editors. Three-dimensional imaging for orthodontics and oral surgery. Chichester, West Sussex(United Kingdom): Wiley-Blackwell; 2010. p. 38–54.

10. Jacobson AF, Ferguson JP. Evaluation of an S.S. White Panorex x-ray machine.Oral Surg Oral Med Oral Pathol 1973;36(3):426–42.

11. Lund TM, Manson-Hing LR. Relations between tooth positions and focal troughsof panoramic machines. Oral Surg Oral Med Oral Pathol 1975;40(2):285–93.

12. Shinozima M, Kohirazawa H, Kubota K, et al. Tomorex (curved rotational tomog-raphy apparatus) in experimental and clinical practice. Oral Surg Oral Med OralPathol 1982;53(1):94–110.

13. Nemtoi A, Czink C, Haba D, et al. Cone beam CT: a current overview of devices.Dentomaxillofac Radiol 2013;42(8):20120443.

14. Available at: http://www.3dorthodontist.com/CBCT_Machines.html. AccessedFebruary 5, 2014.

15. Available at: http://www.conebeam.com/pdf/CBCT_chart_12242012.pdf. Ac-cessed February 5, 2014.

16. National Electrical Manufacturers Association. Digital Imaging and Communica-tions in Medicine (DICOM), PS 3 2011. Rosslyn (VA): National Electrical Manufac-turers Association (NEMA); 2011. Available at: http://medical.nema.org/standard.html. Accessed March 13, 2014.

17. Scarfe WC, Farman AG. What is cone-beam CT and how does it work? Dent ClinNorth Am 2008;52(4):707–30.

18. Miles DA. CBCT: facilitating comprehensive, high-tech diagnostics. CompendContin Educ Dent 2011;32(4):14–5. Available at: http://www.dentalaegis.com/specialissues/2011/12/cbct#sthash.PWm57wLG.dpuf.

19. Andriole KP, Ruckdeschel TG, Flynn MJ, et al. ACR-AAPM-SIIM practice guide-line for digital radiography. J Digit Imaging 2013;26(1):26–37.

20. Bechara B, McMahan CA, Geha H, et al. Evaluation of a cone beam CT artefactreduction algorithm. Dentomaxillofac Radiol 2012;41(5):422–8. http://dx.doi.org/10.1259/dmfr/43691321.

21. Bechara B, Moore WS, McMahan CA, et al. Metal artefact reduction with conebeam CT: an in vitro study. Dentomaxillofac Radiol 2012;41(3):248–53. http://dx.doi.org/10.1259/dmfr/80899839.

22. Kamburoglu K, Kolsuz E, Murat S, et al. Assessment of buccal marginal alveolarperi-implant and periodontal defects using a cone beam CT system with andwithout the application of metal artefact reduction mode. Dentomaxillofac Radiol2013;42(8):20130176.

23. Bechara B, McMahan CA, Moore WS, et al. Cone beam CT scans with andwithout artefact reduction in root fracture detection of endodontically treatedteeth. Dentomaxillofac Radiol 2013;42(5):20120245.