Faculty Disclosures
MR Angiography: Techniques and Pitfalls
Vincent B. Ho, M.D.
Financial Disclosure:
Grant/Research Support: General Electric Medical Systems
Off-Label/Investigational Drug Use:
Dr. Ho will discuss the usage of Gd-chelate contrast agents
for cardiovascular MRI/MRA which is an “off-label” use.
The opinions or assertions contained herein are the private
views of Dr. Ho and are not to be construed as official or
reflecting the views of USUHS or the Dept. of Defense.
Presentation Objectives
At the end of this presentation, the
participant w ill be able:
• To discuss common MR techniques for
vascular evaluation
• To describe common pitfalls of MRA
methods
MRA Techniques
Time-of-Flight MRA
Phase Contrast MRA
Steady State Free Precession MRA
Gd-Enhanced MRA
Vincent B. Ho, M.D., M.B.A.
Professor of Radiology
Uniformed Services University
Bethesda, Maryland
Pros and Cons of MRA
Benef its:
• Does not require iodinated contrast materials
• Fast (especially Gd-enhanced MRA)
• Readily perf ormed on most current scanners
• Does not require ionizing radiation
Disadvantages:
• MRI has contraindications (e.g. pacemakers)
• Stents ty pically cause artifacts on MRI
• Not necessarily “push button”
Time-Of-Flight MRA
Principle:
• In-Flow Effect (a.k.a. “wash-in” phenomenon
or “f low related enhancement”)
Arterial-Venous Segmentation:
• Selective Saturation Band
Concerns:
• Intravoxel Dephasing [Benefit: Flow Jet]
• Saturation Effects (e.g. in-plane flow)
• Pulsatility Artifacts
1
Time-Of-Flight MRA
Time-of-Flight
(“In-Flow” phenomenon)
Principle:
• In-Flow Effect (a.k.a. “wash-in” phenomenon
or “f low related enhancement”)
Arterial-Venous Segmentation:
• Selective Saturation Band
Concerns:
Imaging
Slice
Resultant
Image
Satur ate d Sp ins
• Saturation Effects (e.g. in-plane flow)
• Pulsatility Artifacts
• Intravoxel Dephasing [Benefit: Flow Jet]
Time-of-Flight
Time-of-Flight
(“In-Flow” phenomenon)
(“In-Flow” phenomenon*)
Imaging
Slice
Resultant
Image
Saturation
Band
Imaging
Slice
Resultant
Image
Time-of-Flight
Time-of-Flight
(“In-Flow” phenomenon)
(“In-Flow” phenomenon*)
Imaging
Slice
Resultant
Image
Imaging
Slice
Saturation
Band
Resultant
Image
2
Time-Of-Flight MRA
TOF Pitfall: Spin Saturation
Principle:
• In-Flow Effect (a.k.a. “wash-in” phenomenon
or “f low related enhancement”)
Arterial-Venous Segmentation:
• Selective Saturation Band
Concerns:
Imaging
Slice
• Saturation Effects (e.g. in-plane flow)
• Pulsatility Artifacts
Resultant
Image
Satur ate d Sp ins
• Intravoxel Dephasing [Benefit: Flow Jet]
TOF Pitfall: Spin Saturation
Time-Of-Flight MRA
Principle:
Overestimation of stenosis
• In-Flow Effect (a.k.a. “wash-in” phenomenon
or
or “f low related enhancement”)
mimicing
of vascular
Arterial-Venous
Segmentation:
occlusion
• Selective Saturation Band
Concerns:
Imaging
Slice
Resultant
Image
Satur ate d Sp ins
• Saturation Effects (e.g. in-plane flow)
• Pulsatility Artifacts
• Intravoxel Dephasing [Benefit: Flow Jet]
Phase Contrast MRA
Phase Contrast MRA
Principle:
Principle:
• Phase Shifts
• Phase Shifts
Arterial-Venous Segmentation:
Arterial-Venous Segmentation:
• Velocity and Direction Prescription
• Phase Map Data
• Velocity and Direction Prescription
• Phase Map Data
Concerns:
•
•
•
•
Intravoxel Dephasing [Benefit: Flow Jet]
Pulsatility Artifacts
Velocity and Direction Prescription
Time
Bipolar Flow Encoded Gradients
3
PC Data
Speed
Phase Contrast MRA
Phase Map
Principle:
• Phase Shifts
Arterial-Venous Segmentation:
Dilated
Azygous Vein
Magnitude
Concerns:
•
•
•
•
Pulsatility Artifacts
Intravoxel Dephasing [Benefit: Flow Jet]
Velocity and Direction Prescription
Time
Steady State Free Precession
Phase Contrast MRA
Axia l n on-G d 3 D P C MR A
• Velocity and Direction Prescription
• Phase Map Data
(SSFP, TrueFISP, BalFFE, FIESTA)
Axia l Po st Gd 3 D MR A
Principle:
• T2/T1 ratio (and some “in-flow”)
Ao
Arterial-Venous Segmentation:
Rt. Renal Artery Stenosis
Concerns:
•
•
•
•
Gd i mproves PC MR A!
Pulsatility Artifacts
Intravoxel Dephasing [Benefit: Flow Jet]
Velocity and Direction Prescription
Time
• Anatomic segmentation by operator
Concerns:
• B0 Homogeneity
• Specif ic Absorption Rate (SAR)
• “Contamination” by adjacent structures with
high T2 (eg. biliary tree and gallbladder)
3D SSFP Coronary MRA
RCA
and
LM
Gd-Enhanced 3D MRA
LAD
• Vascular imaging relies on the T1-shortening of
blood by circulating Gd-chelate contrast media
• Imaging during arterial transit (arterial phase) of
the bolus affords preferential arterial illustration
RCA
LCx
Relative Signal Intensity
0.40
0.30
0.25
0.20
0.15
0.10
CNR
0.05
0
Ho VB, et al. M RA C lu b 2 00 3
Gd-enhancement (T1 = 100 msec)
Muscle (T1=600 msec)
Blood (T1=1200 msec)
0.35
0
10 20 30 40 50 60 70 80 90 100
Number of RF pulses
4
Gd-enhanced 3D MRA
Principle:
• Sy nchronization of imaging (center of k-space)
for peak vascular contrast enhancement
Gadolinium-Enhanced MRA
Arterial depiction relies on the synchronization of
imaging with the arterial phase of the Gd bolus
Artery
Arterial-Venous Segmentation:
Signal
Intensity
Vein
• Timing (e.g. arterial phase)
Time
Concerns:
• Patient Preparation
• Timing
• “Ringing Artif act”
Preferential
Arteria l
Enhancement
K-space Data for
Conventional Sequential
3D MRA
?
Center of
k-space
Imag in g D el ay Tim e
Gd-Enhanced MRA: Timing
Gd-Enhanced MRA: Timing
Artery
Artery
Vein
Time
Resultant
Image
Gd-Enhanced MRA: K-space
Timing of Gd-MRA depends on the MRA k-spac e sc heme
Vein
Signal
Intensity
Preferential
Arteria l
Enhancement
Conventional Sequential
Center of
k-space
Imagin g
Volume
Resultant
Image
K-Space Scheme
Conv entional Acquisition (Sequential or Linear)
ky
Artery
k-Space
Scheme
Signal
Intensity
Time
Imagin g
Volume
Time
Vein
Signal
Intensity
1
2
3
4
5
6
7
8
9
10
:
:
kz
Sequential Centric
Elliptical Centric
Partial Fourier Reverse
Sequential
Ho VB, et al . Top Magn Reson Imaging 2001;12:283-299
Ho VB, et al . Top Magn Reson Imaging 2001;12:283-299
5
K-Space Scheme
K-Space Scheme
Conv entional Centric Acquisition (Sequential)
Elliptical Centric Acquisition
ky
ky
:
:
7
5
3
1
2
4
6
8
:
:
K-space Trajectory:
From center to next
closest radial point to
center
kz
Result:
Compact acquisition of
central k-space data
kz
Ho VB, et al . Top Magn Reson Imaging 2001;12:283-299
Ho VB, et al . Top Magn Reson Imaging 2001;12:283-299
K-Space Scheme
K-Space Scheme
Conv entional Partial-Fourier Acquisition (“0.5 NEX”)
Partial-Fourier Reverse Sequential Acquisition
ky
ky
1
2
3
4
5
6
7
8
9
10
:
:
:
:
7
6
5
4
3
2
1
kz
over sca n re gi on f or lo w
spati al fr eq ue ncy estim ati on
over sca n re gi on f or lo w
spati al fr eq ue ncy estim ati on
Ho VB, et al . Top Magn Reson Imaging 2001;12:283-299
Gd-Enhanced MRA: K-space
Timing of Gd-MRA depends on the MRA k-spac e sc heme
Artery
Vein
Time
Signal
Intensity
Preferential
Arteria l
Enhancement
Ho VB, et al . Top Magn Reson Imaging 2001;12:283-299
Gd-MRA: Timing
Timing Scan
Real-Time Triggering
MR Fluoroscopy
MR SmartPrep
Multi-Phase Imaging
Artery
Conventional Sequential
k-Space
Scheme
Center of
k-space
kz
Sequential Centric
Vein
Time
Preferential
Arteria l
Enhancement
Elliptical Centric
Partial Fourier Reverse
Sequential
Ho VB, et al . Top Magn Reson Imaging 2001;12:283-299
?
Imaging Delay Time
6
Test Bolus
Gd-MRA: Timing
Timing Scan
Real-Time Triggering
MR Fluoroscopy
MR SmartPrep
Multi-Phase Imaging
Prescription: Place monitoring slice over
the target vessel
Pulse sequence: axial FMPSPGR, S/I sat,
TR 20 msec, TE min, FA 60 degrees,
31.5 kHz bandwidth, 256x128, 1 NEX
…1 image/1-2 sec
Axial Timing
Run
Injection: 1-2 mL @ 2 mL/sec followed by
25 mL saline flush via MR-compatible
injector
Apply Inferior
and Superior
Sat Bands
Test Bolus for Gd-MRA Timing*
Test Bolus for Gd-MRA Timing*
Td = Time to Arte ri al
Enh an ce me nt
Td = Time to Arte ri al
Enh an ce me nt
Test Bol us
Test Bol us
Tg = Dur atio n of
Arteri al E nh an ce me nt
Di ag nost ic Bo lu s
Ts = Time to
Start of Sca n
Tg = Dur atio n of
Arteri al E nh an ce me nt
Di ag nost ic Bo lu s
Ts = Time to
Start of Sca n
Ta = Data Ac q uis itio n
Ts = Td + Tg/2 - Ta/2
*Earls JP, Rofsky NM, DeCorato DR, et al.
Ts = Td + 2 sec
Radiology 1996;201:705-710
Gd-MRA: Timing
Gd-MRA: Timing
Timing Scan
Real-Time Triggering
MR Fluoroscopy
MR SmartPrep
Multi-Phase Imaging
B
Timing Scan
Real-Time Triggering
MR Fluoroscopy
MR SmartPrep
Multi-Phase Imaging
C
2000
Dete ct Gd
1800
MR Fluoro Trigger
3D MRA w/SENSE
Inte gral of Ec ho Spee d
A
Ta = Data Ac q uis itio n
1600
1400
1200
1000
Gd-MRA (centric)
800
0
10
20
30
40
50
60
Tim e (s ec )
7
Gd-MRA: Timing
Contrast Administration
Timing Scan
Real-Time Triggering
MR Fluoroscopy
MR SmartPrep
Multi-Phase Imaging
• Right antecubital IV (at least 22G)
• Match bolus duration with at least half that of
the imaging time, namely central k-space
• Contrast Dose: 20 to 30 mL @ 2 mL/sec
Hany et al (ISMRM 1998,768) estimated that
need AT LEAST 0.12 mmol/kg
Multi-Phase 3D MRA
• Saline Flush: 30 mL @ 2 mL/sec
Im aging Param eters
• Shortest possible TR and TE
• Optimize spatial resolution vs. time (e.g. 2-3 mm
for aorta; ≤ 2 mm for renal arteries)
Examples: 0.75 FOV, 0.5 NEX, ± 64 kHz
• ZIP (a.k.a. zero filling) … improv es MIP (NOTE:
spatial resolution not really better)
• Centric (or elliptical centric) k-space phase
ordering...partial Fourier using rev erse sequential
• Parallel Imaging (a.k.a. SENSE, ASSET)
Left Sided Venous Injection
Patient Preparation
• Test BH capacity and instruct patient
• Respiratory bellows
• Cardiac gating (cine MR)
• Oxygen and hyperventilation
• Patient positioning
• Coil selection
• Surgical history (e.g. extra-anatomic
shunt, stents)
• Right antecubital IV
Gd-MRA: K-space Scheme
Timing of Gd-MRA depends on the MRA k-spac e sc heme
T2* artifact caus ed by
the high c oncentration
of Gd during the ini tial
“pass” thr oug h the left
brac heoc ephalic vein
Artery
Vein
Time
Sol ution:
Right antec ubi tal
injecti on and
multiphase (delayed
phas e) imaging
Signal
Intensity
Preferential
Arteria l
Enhancement
Conventional Sequential
k-Space
Scheme
Arterial Phase
Delayed Phase
Center of
k-space
Sequential Centric
Elliptical Centric
Partial Fourier Reverse
Sequential
Ho VB, et al . Top Magn Reson Imaging 2001;12:283-299
8
Gd-MRA: K-space Scheme
“Ringing Artifact" on Gd-MRA
Timing of Gd-MRA depends on the MRA k-spac e sc heme
“Ringing Artifact”
Artery
Signal
Intensity
Vein
Time
Preferential
Arteria l
Enhancement
Conventional Sequential
Sequential Centric
k-Space
Scheme
Elliptical Centric
Partial Fourier Reverse
Sequential
Center of
k-space
Acknowledgements
Bill Corse
Julianna Czum
Thomas Foo
Kai Yiu Ho
Maureen Hood
Scott Pereles
Martin Prince
Barry Stein
9