Handout 1

Answers for life. Unrestricted © Siemens Healthcare GmbH 2015 All rights reserved.

Trends in MR Angiography Shivraman Giri, Ph.D.

Siemens Healthcare, Chicago, IL, USA

Disclosures

Full-time employee of Siemens Medical Solutions

Acknowledgements

NorthShore University Health Systems

Robert R. Edelman, MD

Northwestern University

James Carr, MD

Michael Markl, PhD

Ohio State University

Orlando P Simonetti, PhD

Siemens

Gerhard Laub, PhD

Angio procedures in the last 15 years: MRA

Source: IMV Report

Angio procedures in the last 15 years: MRA vs CTA

Source: IMV Report

Many MR Angiography applications

Handout 2

Outline

• Review of basic MRA principles

Contrast enhanced

Non-enhanced

• Building blocks of pulse sequences

• Advanced acceleration techniques

• Clinical MRA pulse sequences

Fundamental requirements of MR Angiography

1. Maximize the signal in desired vessel (eg. artery)

2. Minimize the signal in

• Background muscle

• Undesired vessel (eg. vein)

• Fat

Properties of arteries that can be used for contrast

Background muscle

T1 ~1000 ms

T2 ~ 40-60 ms

No motion (relatively static)

Fat

T1 ~250 ms

T2 ~ 70 ms

No motion (relatively static)

Arterial blood

T1 ~ 1200 ms

T2 ~ 250 ms

Pulsatile flow

Venous blood

T1 ~ 1200 ms

T2 ~ 250 ms

Slow constant flow

T1W imaging?

Background muscle

T1 ~1000 ms

Fat

T1 ~250 ms

Arterial blood

T1 ~ 1200 ms

Venous blood

T1 ~ 1200 ms

In presence of contrast agent, blood

T1 ~ 50-100 ms

MRA: different techniques

Using T1 shortening

contrast agents

• Appropriate timing to avoid venous

T1 shortening

• Still need to account for fat

T2W imaging?

Background muscle

T2 ~ 40-60 ms

Fat

T2 ~ 70 ms

Arterial blood

T2 ~ 250 ms

Venous blood

T2 ~ 250 ms

Handout 3

MRA: different techniques

Using T1 shortening

contrast agents

• Appropriate timing to avoid venous

T1 shortening

• Still need to account for fat

Non-contrast techniques

Flow independent

• T2W

• Veins visible

Properties of arteries that can be used for contrast

Background muscle

No motion (relatively static)

Fat

No motion (relatively static)

Arterial blood

Pulsatile flow

Venous blood

Slow constant flow

MRA: different techniques

Using T1 shortening

contrast agents

• Appropriate timing to avoid venous

T1 shortening

• Still need to account for fat

Non-contrast techniques

Flow independent

• T2W

• Veins visible

Flow dependent

• Spin echo washout

• GRE enhancement

• Phase contrast MRI

Time of flight effects in MRI

Spin Echo (SE)

Gradient Echo (GRE)

a

TR

TE

GRE-Signal

//

90°

180°

SE-Signal

TE

TR

//

Spin-echo washout effect

Spin Echo: Out-Flow / Wash-out Effect

• Spins move out of slice prior to 180o refocusing

Flow void, dark blood

• Depends on TE, slice thickness & velocity

90°

180°

SE-Signal

TE

TR

//

90°

180°

SE-Signal

TE

TR

//

Blood

Flow, vBlood

slice

Ds Complete outflow: vBlood TE/2 > Ds

90°

Distance = vBlood TE/2

180°

90°

Question

T1-Spin Echo - Short TE

Arterial blood: complete out-flow dark blood

Venous blood: Slow flow, no out-flow bright blood

Blood

Flow

slice

Ds

90°

Distance

= vBlood TE/2

Distance

= vBlood TE/2

180°

Different (dark / bright)

vessel signal ?

Handout 4

Question

What happens to arterial signal in

1. Systolic phase (fast blood flow)

• High or low?

2. Diastolic phase (slow blood flow)

• High or low?

Answer: low

Answer: high

Gated subtracted techniques

Lim R P et al. Radiology doi:10.1148/radiol.12120859

Comic relief

Albert Mackowski, Stanford University

Turn the liver Excitation

Water gushing in bowl Signal

Decays fast with “T2” (~10 seconds)

Wait for the flush-tank to fill again

“T1 recovery” (~60 seconds)

Takes longer than “T2” decay

Toilet flush analogy to MRI

What happens if you turn the liver again within 10 seconds?

GRE in-flow enhancement effect

Flow

Static spins saturated

by repeated excitation

Flow

a

TR

TE

GRE-Signal

//

a

TR

TE

GRE-Signal

//

In-flow of

'fresh' spins

Gradient Echo: In-Flow Effect

vBlood TR

TOF sequence

arterial

venous

venous arterial

QISS sequence

Handout 5

Phase contrast MRI

MR Signal = Vector

Signal Phase ~ Flow Measurement of blood flow

Phase contrast MRI

Flow-Encoding

Phase ~ local velocity &

bipolar gradient

But: unknown background phase

Reference-Measurement & Subtraction

- Flow in 1 direction - 2 measurements

Ph

ase

Static Spins

Ph

ase

Bipolar

Gradient

Moving Spins

f ~ v

G

t

f ~ phase

static spins

phase

moving spins

+

Phase contrast MRI

Re

fere

nce

Flo

w s

en

sitiv

e

• Velocities encoded in phase difference image

Phase difference, flow Magnitude

x

y

Flow, z

x

y

Flow, z

Summary of different MRA techniques

Using T1 shortening

contrast agents

• Appropriate timing to avoid venous

T1 shortening

• Still need to account for fat

Non-contrast techniques

Flow independent

• T2W

• Veins visible

Flow dependent

• Spin echo washout

• GRE enhancement

• Phase contrast MRI

MR pulse sequence menu

Magnetization

preparation

o Inversion Recovery

o Saturation Recovery

oDark Blood

o STIR

o Fat Suppression

o T2 Preparation

oDiffusion Weighting

o Tagging

oDENSE

oMagnetization

Transfer

o Velocity Encoding

Echo formation

o SE

Turbo SE

SPACE

oGRE

o SSFP

o EPI

SE-EPI

GRE-EPI

o FID

K-space

trajectory

oCartesian

Linear

Centric

o Spiral

oRadial

o BLADE etc.

o TWIST, Keyhole etc.

K-space

segmentation

oNon-segmented

o Segmented

o Single-shot

Image

Reconstruction

o Partial fourier

o Parallel Imaging

SENSE

GRAPPA

TSENSE

TGRAPPA

CAIPIRINHA

o k-t methods

oCompressed sensing

Advanced acceleration techniques

Courtesy of Dr. Peter Killman, NIH

Parallel Imaging SENSE

GRAPPA CAIPIRINHA

exploit coil sensitivities

Handout 6

Advanced acceleration techniques

Courtesy of Dr. Peter Killman, NIH

Temporal Viewshare/sliding block

-TWIST UNFOLD/filtering BRISK Interpolation

exploit coil sensitivities

exploit frame-to-frame temporal

correlation

k-t BLAST

TSENSE TGRAPPA

Parallel Imaging SENSE

GRAPPA CAIPIRINHA

Advanced acceleration techniques

Courtesy of Dr. Peter Killman, NIH

Modeling BLAST

RIGR

Temporal Viewshare/sliding block

-TWIST UNFOLD/filtering BRISK Interpolation

exploit a’priori

exploit coil sensitivities

exploit frame-to-frame temporal

correlation

k-t BLAST

k-t SENSE x-f Choice

TSENSE TGRAPPA

Parallel Imaging SENSE

GRAPPA CAIPIRINHA

Basic principles of Contrast-Enhanced MRA

Injection of contrast-agent

T1 shortening of blood

T1-weighted imaging protocols

Fast protocols to get data from arterial phase only

How it works

Low signal of background tissue

– saturation of spins with long T1

High signal intensity of blood

– no saturation of blood due to T1-shortening

CE-MRA pulse sequence

Magnetization

preparation

o Inversion Recovery

o Saturation Recovery

oDark Blood

o STIR

o Fat Suppression

o T2 Preparation

oDiffusion Weighting

o Tagging

oDENSE

oMagnetization

Transfer

o Velocity Encoding

Echo formation

o SE

Turbo SE

SPACE

oGRE

o SSFP

o EPI

SE-EPI

GRE-EPI

o FID

K-space

trajectory

oCartesian

Linear

Centric

o Spiral

oRadial

o BLADE etc.

o TWIST, Keyhole etc.

K-space

segmentation

oNon-segmented

o Segmented

o Single-shot

Image

Reconstruction

o Partial fourier

o Parallel Imaging

SENSE

GRAPPA

TSENSE

TGRAPPA

CAIPIRINHA

o k-t methods

oCompressed sensing

Contrast Agent Options for CEMRA

Standard Extra-Cellular Agents First pass

Double dose for steady-state imaging of veins

High Relaxivity Agents (Gadobenate, Gadobutrol) Higher SNR or decreased contrast agent dose

Blood Pool Agents (Gadofosveset, Ferumoxytol) First pass

Steady-state imaging for high resolution imaging of arteries, venous imaging

Handout 7

FDA approved MRI contrast agents

Adapted from Runge VM. Top Magn Reson Imaging. 2001;12:309.

Blood pool agents

http://www.ablavar.com/image-gallery.html

First pass Steady-state

• First pass => separates arteries and veins, short scan time allows breath-holding

• Steady-state => no time constraint on scan time, higher spatial resolution, veins well seen

(but may overlap the arteries)

Timing considerations in CE-MRA

rel.

sig

nal

en

han

cem

en

t

time

venous phase

arterial phase acquisition window

Timing considerations in CE-MRA

k=0

Early

Correct

Test Bolus

time

venous phase

arterial phase contrast injection

(~2cc)

TurboFLASH, 1 image/sec

Test Bolus

4 sec 5 sec 6 sec 7 sec 8 sec 9 sec 3 sec

11 sec 12 sec 13 sec 14 sec 15 sec 16 sec 10 sec

Handout 8

Alternatively: "Care-Bolus"

3D

contrast Injection

(full dose)

... ... Image reconstruction

2D 2D 2D 2D 2D 2D 2D 2D 2D

Circumvent timing issues with fast time-resolved CE MRA

Contrast injection (~5 cc at 2 cc/sec)

Avoids issues with timing and venous

enhancement

Time-resolved CE-MRA pulse sequence

Magnetization

preparation

o Inversion Recovery

o Saturation Recovery

oDark Blood

o STIR

o Fat Suppression

o T2 Preparation

oDiffusion Weighting

o Tagging

oDENSE

oMagnetization

Transfer

o Velocity Encoding

Echo formation

o SE

Turbo SE

SPACE

oGRE

o SSFP

o EPI

SE-EPI

GRE-EPI

o FID

K-space

trajectory

oCartesian

Linear

Centric

o Spiral

oRadial

o BLADE etc.

o TWIST, Keyhole etc.

K-space

segmentation

oNon-segmented

o Segmented

o Single-shot

Image

Reconstruction

o Partial fourier

o Parallel Imaging

SENSE

GRAPPA

TSENSE

TGRAPPA

CAIPIRINHA

o k-t methods

oCompressed sensing

Aortic dissection

CTA Time-resolved CEMRA

Peripheral run-offs

Meas

4

Meas

2

Meas

3

Meas

1

Meas

5

time wait for contrast arrival

Meas

6

po

sit

ion

Peripheral run-offs

3 1 2

Handout 9

Any questions on CE-MRA Motivations for Non-enhanced MRA

• Provide alternative to CEMRA in patients with poor renal function

(avoid NSF risk)

• Reduce cost (e.g. eliminate need for contrast agents, disposables, and

point-of-service renal function testing)

=> >$100 - $200/patient!

• Provide backup in case of technical failure of CEMRA

• Reduce need for specialized technologist expertise

(e.g. required for peripheral CEMRA)

MRA: different techniques

Using T1 shortening

contrast agents

• Appropriate timing to avoid venous

T1 shortening

• Still need to account for fat

Non-contrast techniques

Flow independent

• T2W

• Veins visible

Flow dependent

• Spin echo washout

• GRE enhancement

• Phase contrast MRI

Which non-enhanced technique to use?

Factors determining the choice of sequence

• Physiologic motion (breathing/cardiac)?

• Surrounding fat?

• Vessel orientation? (vertical/horizontal/tortuous)

• Arteries and veins in close proximity

TOF

3D for brain

2D for carotids

Flow-

independent

3D inflow

enhancement

Fast spin echo

with subtraction

Recent: QISS

2D inflow

enhancement

Head/Neck NCE-MRA

Magnetization

preparation

o Inversion Recovery

o Saturation Recovery

oDark Blood

o STIR

o Fat Suppression

o T2 Preparation

oDiffusion Weighting

o Tagging

oDENSE

oMagnetization

Transfer

o Velocity Encoding

Echo formation

o SE

Turbo SE

SPACE

oGRE

o SSFP

o EPI

SE-EPI

GRE-EPI

o FID

K-space

trajectory

oCartesian

Linear

Centric

o Spiral

oRadial

o BLADE etc.

o TWIST, Keyhole etc.

K-space

segmentation

oNon-segmented

o Segmented

o Single-shot

Image

Reconstruction

o Partial fourier

o Parallel Imaging

SENSE

GRAPPA

TSENSE

TGRAPPA

CAIPIRINHA

o k-t methods

oCompressed sensing

Head/Neck NCE-MRA

3D imaging volume

Pre-sat region

3D-TOF

• volumetric coverage

• high resolution

Arterial

Flow

Venous

Flow

Handout 10

Head/Neck NCE-MRA

3D imaging volume

Pre-sat region

3D-TOF Challenges

• Saturation of blood by repeated slab excitation

Reduced blood-tissue contrast within 3D volume

Arterial

Flow

Venous

Flow

Blood signal

Head/Neck NCE-MRA

3D imaging volume

Arterial

Flow

Arterial

Flow

Slab 1 Slab 2 Slab 3 Slab 4

3D-TOF Challenges

• Compensate for blood saturation

Multiple smaller 3D volumes

Head/Neck NCE-MRA

Multiple Slab 3D-TOF

• Minor saturation effects

• Signal discontinuities at slab borders

Thoracic – NCE-MRA

Factors determining the choice of sequence

• Physiologic motion (breathing/cardiac)?

• Surrounding fat?

• Vessel orientation? (vertical/horizontal/tortuous)

• Arteries and veins in close proximity

TOF

3D for brain

2D for carotids

Flow-

independent

3D inflow

enhancement

Fast spin echo

with subtraction

Recent: QISS

2D inflow

enhancement

Thoracic – NCE-MRA

Magnetization

preparation

o Inversion Recovery

o Saturation Recovery

oDark Blood

o STIR

o Fat Suppression

o T2 Preparation

oDiffusion Weighting

o Tagging

oDENSE

oMagnetization

Transfer

o Velocity Encoding

Echo formation

o SE

Turbo SE

SPACE

oGRE

o SSFP

o EPI

SE-EPI

GRE-EPI

o FID

K-space

trajectory

oCartesian

Linear

Centric

o Spiral

oRadial

o BLADE etc.

o TWIST, Keyhole etc.

K-space

segmentation

oNon-segmented

o Segmented

o Single-shot

Image

Reconstruction

o Partial fourier

o Parallel Imaging

SENSE

GRAPPA

TSENSE

TGRAPPA

CAIPIRINHA

o k-t methods

oCompressed sensing

Thoracic – NCE-MRA

Coronary artery imaging MRA of great vessels

Cardiac and respiratory gating used

Handout 11

Abdominal – NCE-MRA

Factors determining the choice of sequence

• Physiologic motion (breathing/cardiac)?

• Surrounding fat?

• Vessel orientation? (vertical/horizontal/tortuous)

• Arteries and veins in close proximity

TOF

3D for brain

2D for carotids

Flow-

independent

3D inflow

enhancement

Fast spin echo

with subtraction

Recent: QISS

2D inflow

enhancement

Abdominal – NCE-MRA

Magnetization

preparation

o Inversion Recovery

o Saturation Recovery

oDark Blood

o STIR

o Fat Suppression

o T2 Preparation

oDiffusion Weighting

o Tagging

oDENSE

oMagnetization

Transfer

o Velocity Encoding

Echo formation

o SE

Turbo SE

SPACE

oGRE

o SSFP

o EPI

SE-EPI

GRE-EPI

o FID

K-space

trajectory

oCartesian

Linear

Centric

o Spiral

oRadial

o BLADE etc.

o TWIST, Keyhole etc.

K-space

segmentation

oNon-segmented

o Segmented

o Single-shot

Image

Reconstruction

o Partial fourier

o Parallel Imaging

SENSE

GRAPPA

TSENSE

TGRAPPA

CAIPIRINHA

o k-t methods

oCompressed sensing

Abdominal NCE-MRA: IR-Prepared 3D bSSFP MRA of Renal Arteries

180

degree

Abdominal NCE-MRA: IR-Prepared 3D bSSFP MRA of Renal Arteries

Resp. Signal

TrueFISP Readout

FS IR1

TI1 = 1350 ms

Sequence Timing IR2

TI2 = 800 ms

--> IR1

--> IR2

Scan Planning Signal Characteristics

Peripheral NCE-MRA

Factors determining the choice of sequence

• Physiologic motion (breathing/cardiac)?

• Surrounding fat?

• Vessel orientation? (vertical/horizontal/tortuous)

• Arteries and veins in close proximity

TOF

3D for brain

2D for carotids

Flow-

independent

3D inflow

enhancement

Fast spin echo

with subtraction

Recent: QISS

2D inflow

enhancement

Peripheral NCE-MRA

Magnetization

preparation

o Inversion Recovery

o Saturation Recovery

oDark Blood

o STIR

o Fat Suppression

o T2 Preparation

oDiffusion Weighting

o Tagging

oDENSE

oMagnetization

Transfer

o Velocity Encoding

Echo formation

o SE

Turbo SE

SPACE

oGRE

o SSFP

o EPI

SE-EPI

GRE-EPI

o FID

K-space

trajectory

oCartesian

Linear

Centric

o Spiral

oRadial

o BLADE etc.

o TWIST, Keyhole etc.

K-space

segmentation

oNon-segmented

o Segmented

o Single-shot

Image

Reconstruction

o Partial fourier

o Parallel Imaging

SENSE

GRAPPA

TSENSE

TGRAPPA

CAIPIRINHA

o k-t methods

oCompressed sensing

Handout 12

Peripheral NCE-MRA

Gated fast spin echo

Peripheral NCE-MRA

V A

slow flow

V A

fast flow

- = V A

subtraction, MIP

Peripheral NCE-MRA – Diseased vessel

Time after R-wave

Sig

nal

inte

nsity

Systole Diastole

Vein

Diseased artery

200 400 600 800 (ms)

A

Systole

A

Diastole

A

Subtraction

Peripheral NCE-MRA

66-year-old male smoker with right

sided claudication

Lack of arterial visualization in the right

calf with Subtractive FSE MRA.

CE MRA Subtractive FSE

stent

Peripheral NCE-MRA using QISS

Magnetization

preparation

o Inversion Recovery

o Saturation Recovery

oDark Blood

o STIR

o Fat Suppression

o T2 Preparation

oDiffusion Weighting

o Tagging

oDENSE

oMagnetization

Transfer

o Velocity Encoding

Echo formation

o SE

Turbo SE

SPACE

oGRE

o SSFP

o EPI

SE-EPI

GRE-EPI

o FID

K-space

trajectory

oCartesian

Linear

Centric

o Spiral

oRadial

o BLADE etc.

o TWIST, Keyhole etc.

K-space

segmentation

oNon-segmented

o Segmented

o Single-shot

Image

Reconstruction

o Partial fourier

o Parallel Imaging

SENSE

GRAPPA

TSENSE

TGRAPPA

CAIPIRINHA

o k-t methods

oCompressed sensing

Peripheral NCE-MRA using QISS

Inferior

Venous Sat

In-Plane

Sat

Quiescent

Interval

≈300 ms

(in-flow of

unsaturated blood)

ky = 0

2D trueFISP – Single Shot

Trigger

Delay

≈100 ms

ECG

R-wave

Fat

Sat

Partial Four

QISS Quiescent Inflow Slice- Selective

Handout 13

Peripheral NCE-MRA using QISS

1

2

3

4

5

6

7

8

9

Peripheral NCE-MRA using QISS

CE MRA Subtractive FSE QISS

stent

66-year-old male smoker with right

sided claudication

Lack of arterial visualization in the right

calf with Subtractive FSE MRA.

Peripheral NCE-MRA using QISS

Subtractive

TSE Subtractive

TSE

Sensitivity to motion

Peripheral NCE-MRA using QISS

Easy workflow

No contrast

No subtraction

Set‘n’Go

Output: all images - one series (CT like)

Peripheral NCE-MRA using QISS

Case courtesy of Drs. Joseph U. Schoepf and Akos Varga-Szemes, Medical Univ of South Carolina, SC, USA

Better than CTA in some cases with calcification

Summary

• There are many flavors of MRA, encompassing contrast-enhanced and

nonenhanced techniques

• CEMRA remains the dominant technique for MRA, with ongoing improvements

in spatial and temporal resolution as well as novel contrast agents

• Newer nonenhanced MRA techniques permit screening for vascular disease with

no risk and high accuracy, eliminating need for (and cost of) contrast agents in

many cases and providing backup in case of technical failure of CEMRA

• Flexibility and accuracy of MRA, along with minimal to no risk, should allow it

to compete successfully with CTA