Arterial Spin Labelling Neuroradiological Clinical Practice Xavier Golay, Ph.D.

Arterial Spin Labelling
Neuroradiological Clinical
Practice
Xavier Golay, Ph.D.
Chair, MR Neurophysics and Translational Neuroscience
UCL Institute of Neurology
Queen Square, London
Disclosure
• Part of the work presented here was done in
collaboration with Philips Medical Systems
(now Philips Healthcare)
Layout
• Introduction and Definitions
• Differences between DSC and ASL
• Clinical ASL Examples
-Atherosclerotic Diseases
-Stroke
-AVMs
-Pediatrics
-Dementia
-Tumours
• Conclusions
Definitions
• The Cerebral Perfusion can be defined as the
steady state delivery rate of blood to the tissue
capillary bed
• The main parameter measured is the Cerebral
Blood Flow (CBF), which is the rate of delivery of
blood to the capillary bed
• Its measurement relies on the Indicator Dilution
Theory
Quantitative Perfusion Measurement
Kety-Schmidt Method (1945)
N2O
MRI-based Perfusion Imaging
• Free-diffusible Tracers (2H, 19F, 17O, 129Xe)
• Arterial Spin Labeling (ASL)
- Continuous ASL
- Pulsed ASL
• Dynamic Susceptibility Contrast (DSC) Perfusion
Imaging (Gd-based)
Theoretical Differences
arteries/arterioles
Gadolinium
Labeled Water
capillaries
venules/veins
Theoretical Differences
• Gadolinium is an intravascular tracer (at least
without disruption of BBB)
- Assessment of the complete vascular compartment size
(I.e. Cerebral Blood Volume or CBV) is possible
• ASL can be considered as a free diffusible tracer
- Assessment of the complete vascular compartment size
(CBV) is NOT possible
Arterial Spin Labeling Basics
Control
M
Label
Tracer = Labeled Magnetization
Pulsed Arterial Spin Labeling
• Pulsed Arterial Spin Labeling
-STAR (Signal Targeting by Alternating Radio-frequency pulses)
Label: 180°
Edelman & Chen. MRM 40(6): 800 (1998)
Golay et al., MRM, 53:15 (2005)
Pulsed Arterial Spin Labeling
• Pulsed Arterial Spin Labeling
-STAR (Signal Targeting by Alternating Radio-frequency pulses)
Control: 180° + 180°
Edelman & Chen. MRM 40(6): 800 (1998)
Golay et al., MRM, 53:15 (2005)
ASL: Model-Free CBF quantification
Tissue signal
Arterial signal
Tissue response
t
M (t )
2 M 0,b f
c( ) r (t
) m(t
0
Meier & Zierler, J Appl Physiol, 6:731(1954)
CT: Gobbel et al, Phys Med Biol, 39:1133(1994)
MRI: Ostergaard et al, MRM, 36:715(1996)
ASL: Buxton et al., MRM, 40:383(1998)
)d
ASL: Model-Free Perfusion Imaging
50
350
650
950
1250
Non crushed data
Crushed data (Venc = 3cm/s)
Arterial Input Function
Petersen et al., MRM, 55: 219 (2006)
1550
1850
2150
2450 [ms]
ASL: Model-Free Perfusion Imaging
• Deconvolution:
-SVD or other regularization method
-Circular SVD* chosen here
-Maximum of Residue function equals the perfusion!
Max = CBF = 75 ml/min/100g
Model !
*Ona Wu, MRM, 50:164(2003)
ASL Parametric Maps
Reason: Left Cortical CVA & Hypertension
ASL Parametric Maps
ASL Parametric Maps: CBF
70 ml/min/100g
0
ASL Parametric Maps: aBV / CBVa
3 ml/100g
0
Advantages of DSC over ASL
• Very robust method to depict perfusion deficits in
numerous diseases
• Rapid (~1min)
• Uses standard sequences (GRE/SE-EPI)
• High SNR
• Capable of measuring delays in arrival times of
several seconds
• Automated processing available
Advantages of ASL over DSC
• Does not require contrast (NFS, children, …)
• Can be repeated (while Gd is dose-limited)
• Provides absolute quantification of CBF
• Has a better temporal resolution
• Can be made insensitive to large vessel artifacts
• Can be combined with selective excitation
Atherosclerotic Diseases
“Borderzone sign” on ASL
•
In patients with normal PWI, more
information may be available with ASL
• 43% show bilateral dropout of ASL signal in
watershed regions
• Serpiginous high ASL in cortical sulci
• “Borderzone sign”
•
Additional imaging findings related to slow
flow.
– Collaterals, aneurysms, fistulas
Slide courtesy of G. Zaharchuk
Normal
Mild
Moderate
Severe
Zaharchuk et al.
Radiology 2009
Identification of Pathological Status
in a Patient with Carotid Stenosis
Uchihashi et al, ISMRM 09
Comparison of CBF values obtained
by ASL & SPECT
in Patients with Carotid Stenosis
Uchihashi et al, ISMRM 09
Comparison of CBF values obtained
by ASL & SPECT
in Patients with Carotid Stenosis
Uchihashi et al, ISMRM 09
Territorial (or selective) ASL
Selective labeling: Left ICA
Hendrikse et al. Stroke 35 (2004)
Territorial ASL
Selective labeling: Right ICA
Hendrikse et al. Stroke 35 (2004)
Territorial ASL
Selective labeling: Vertebral arteries (POST)
Hendrikse et al. Stroke 35 (2004)
ASL: Labeling T-ASL
Hendrikse et al. Stroke 35 (2004)
Patient with left ICA Occlusion
TASL vs. DSA
Chng et al. Stroke 39 (2008)
TASL vs. DSA
Chng et al. Stroke 39 (2008)
TASL vs. DSA
• Significant Contingency between RPI & DSA
(V=0.53, C=0.67, p<0.0001)
• “Substantial Agreement” between methods when compared
using a weighted kappa test (k=0.70 flow, k=0.72 collaterals)
Chng et al. Stroke 39 (2008)
Stroke
Singapore Stroke study: Scan Protocol
• T2W + DWI
• 3D TOF MRA
• Two ASL scan
- Global perfusion (non-selective, label all vessels at the same time)
- Territorial perfusion (selectively label individual vascular territory)
• Dynamic Susceptibility Contrast
- Single dose (2 ml/kg Gadolinium)
- Power Injector 5/ml/s
Patient 1: Complete work-up
Hendrikse et al. Stroke, in press (2009)
Patient 2: Complete work-up
DWI
PWI
(CBF)
PWI
(MTT)
ASL
(CBF)
ASL
(AT)
ASL
(TASL)
ASL versus PWI
ASL (left) / PWI (right)
Non-affected
hemisphere
(N=87)
ASL versus PWI
ASL-CBF vs. PWI-CBF (Ratio = ipsi/cont.)
Within infarct (>8cc)
(>20cc)
r=0.55, p=0.0015
Peripheral (8mm)
r=0.54, p=0.0014
Preliminary Conclusion on Singapore Stroke Study
• Success rate: 91.7% in 160 patients
• Changed classification in 11% of patients with cortical /
borderzone infarcts
• Added perfusion information in stroke
- Risk management of large vessel disease
• DSC and ASL provide similar, yet complimentary information
Hendrikse et al, Stroke 40 (2009)
Clinical evaluation
Stroke patient
Slide courtesy from M. Guenther
ASL can visualize collateral flow
74 year-old man with right facial
droop
Serpiginous high ASL
signal surrounding the
infarct represents slow
flow in collateral vessels
Bolus PWI Tmax
ASL without vessel
suppression
ASL with vessel
suppression
Slide courtesy from G. Zaharchuk
Arterio-Venous Malformations
49 F, known AVM
T2
ASL CBF
PWI CBF
Slide courtesy from G. Zaharchuk
4D-MRA & selective ASL
45 y/o patient, grade II AVM, 2 feeding arteries
AVM
left ICA
left ICA
Functional crossfilling documented by selective ASL
Slide courtesy from W. Willinek
Paediatrics
CASL: Sickle Cell Patient
250 ml/min/100g 255 ml/min/100g
0 ml/min/100g
K. Oguz, et al. Radiology (2003)
0 ml/min/100g
CASL: Sickle Cell Patient 2
255 ml/min/100g
0 ml/min/100g
K. Oguz, et al. Radiology (2003)
Sickle Cell : Results
K. Oguz, et al. Radiology (2003)
Dementia and Aging
ASL application
Age dependency of perfusion
15 years
56 years
ASL
CBF
GREDSC
rCBF
Slide courtesy from M. Guenther
Warmuth, C; Günther, M; Zimmer, C.: Radiology 2003, Vol. 228, pp. 523-532
Alzheimer’s Disease Patient
Collaboration with Dr. Christopher Chen, NUS, Singapore
Vascular Dementia Patient
Collaboration with Dr. Christopher Chen, NUS, Singapore
ASL in Dementia (Chinese population)
Brain Perfusion Reduction (Normal vs. Demented)
Mak et al. ISMRM (2010)
ASL in Dementia
Brain Perfusion Reduction (Normal vs. Demented)
Mild Cognitive Impairment
Johnson et al. Radiology, 234:851–859 (2005)
Alzheimer’s Disease
Tumours
Cerebral blood flow and perfusion
Comparison DCE-ASL
67y, female, brain metatasis, before and after radiotherapy
T2w
T1w
ASL
DSC
before
after
Slide courtesy from M. Guenther
M.-A.Weber, A. Kroll, M.Günther et al, Radiologe 2004 · 44:164–173
Cerebral blood flow and perfusion
Comparison DCE-ASL
59y, female, meningioma
Slide courtesy from M. Guenther
ASL
DSC
T1w
T2w
M.-A.Weber, A. Kroll, M.Günther et al, Radiologe 2004 · 44:164–173
Pathologies
• Routine clinical work
ASL
• Pharmacological Studies
ASL
• Stroke
DSC / ASL
• Carotid Artery Stenosis / TIA
DSC / ASL
• Moya-Moya
DSC / ASL
• Dementia
ASL
• Brain Tumors:
ASL / DCE
- Primary brain lesions (Gliomas, etc…)
- Meningiomas
- Metastases
• Children
ASL
Conclusions
• In the last years, ASL managed to catch-up with DSC
-Recognition of Arterial Transit Time Artifacts
-Providing Quantification (CBF, aBV, ATT)
• DSC is still the preferred method in pathologies
involving very large arterial delays (Moya-moya,
Carotid stenosis, …)
• ASL is an add-on in patients where no perfusion
measurement was present (dementia, epilepsy, …)
• ASL should be the method of choice in pediatrics and
renal failure patients
Acknowledgments
• Slides: Yoshito Uchihashi, Greg Zaharchuk,
Winfried Willinek, Matthias Guenther
• Staff from NNI: Esben Petersen, Lynn Ho, Ivan
Zimine, Tchoyoson Lim, Yih Yian Sitoh, Francis Hui
• National Medical Research Council (NMRC),
Singapore, NMRC/0919/2004
• Singhealth Foundation, Singapore, NHGARPR/04012
• Philips Medical Systems
ASL Network
www.asl-network.org/