The Effect of Bolus Length and Dispersion on Arterial Spin Labeling Flow Quantification
Esben Thade Petersen¹, Xavier Golay², T QUASAR Reproducibility Study^3
1Clinical Imaging Research Centre (CIRC), Singapore, Singapore; ²UCL Institute of Neurology, London, United Kingdom; ³28 Centers

Bolus duration and dispersion is often assumed when quantifying flow using ASL. We evaluated their impact on CBF, based on data from 284 healthy subjects (28 sites). The length and dispersion was fitted from multiple arterial-input-functions obtained from data acquired at multiple time-points. Although QUIPSS-II bolus definition (0.64s) was applied, the majority had shorter boluses, compromising the precision of ASL. Furthermore, a considerable correlation (0.63, p<0.001) between average bolus-length and CBF from the sites, suggest that part of site differences relates to the bolus duration. Normal Gaussian dispersion ranges from 0.05-0.15s potentially introducing large quantification errors across the brain.

Determination of Spin Compartment in ASL Signal Using TRUST-MRI
Peiying Liu Wang¹, Jinsoo Uh¹, Hanzhang Lu^1
1Advanced Imaging Research Center, University of Texas Southwestern Medical Center, Dallas, TX, United States

Although ASL has been widely used for measurement of CBF, we do not know which compartment the labeled spins are located at the time of detection. Here we used the T2 value of the labeled spins to probe whether the detected ASL signal is located in artery, tissue or even vein. Our data suggest that, at typical delay time of 1.5 seconds, most of the detected spins in gray matter are already in the tissue space. For white matter, however, the spins are still virtually all in arteries.

Depression of Cortical Gray Matter CMRO₂ in Awake Humans During Hypercapnia
Divya S. Bolar^1,2, Bruce R. Rosen^1,2, Karleyton C. Evans^1,3, A Gregory Sorensen^1,2, Elfar Adalsteinsson^1,2
1HST/MGH/MIT Martinos Center for Biomedical Imaging, Charlestown, MA, United States; ²Harvard-MIT Division of Health Sciences & Technology, Cambridge, MA, United States; ³Department of Psychiatry, Massachusetts General Hospital, Boston, MA, United States

Hypercapnia induced by CO₂ inhalation causes a robust increase in cerebral blood flow. Far less understood are the effects of CO₂ on neuronal activity and cellular metabolism.  In this study, a recently developed method called QUantiative Imaging of the eXtraction of Oxygen and TIssue Consumption (QUIXOTIC) was used evaluate the hypercapnic CMRO₂ response in cortical gray matter of awake humans.  We report a statistically significant decrease of 25.3% in cortical CMRO₂ (p = 0.036), from normocapnia to hypercapnia.  To our knowledge, this is the first time cortical GM CMRO₂ response to hypercapnia has been assessed.

3D-EPI ASL at Ultra High Field
Emma Louise Hall¹, Penny A. Gowland¹, Susan T. Francis^1
1Sir Peter Mansfield Magnetic Resonance Centre, University of Nottingham, Nottingham, Nottinghamshire, United Kingdom

3D acquisitions are advantageous to ASL to eliminate slice dependent variation in signal. Here we show the feasibility of 3D-EPI arterial spin labelling (ASL) at 7T. Using SENSE acceleration in two directions the shot length can be significantly reduced allowing improved spatial coverage or spatial resolution to be achieved. 3D-EPI ASL is shown to benefit from increased signal-to-noise ratio and overcome SAR limits reached when using  2D-EPI ASL at 7T.Whole head (20 slice) 2x2x3mm³ 3D-EPI perfusion images can be acquired in 5 minutes.

Whole Brain Pseudo Continuous ASL at 7T Using a Single Coil for Imaging and Labeling
Wouter M. Teeuwisse¹, Andrew Webb¹, Matthias J.P. van Osch^1
1C.J.Gorter Center, Radiology, Leiden University Medical Center, Leiden, Netherlands

In this study, whole brain pseudo continuous ASL (pCASL) is implemented at 7T, using the same RF coil for labeling and imaging. The magnitude of B0 inhomogeneities, RF penetration and f0-offsets were measured. For optimal labeling, B0 changes along the vessels were compensated by adjusting the average labeling gradient. A subject-specific frequency offset for the label pulses was calculated and implemented as was the incorporation of high dielectric material placed around the head and neck for higher B1 delivery in the neck. After implementing all of these improvements whole brain pCASL was successfully performed at 7T.

Optimizing the Inversion Efficiency of Pseudo-Continuous ASL Pulse Sequence Using B0 Field Map Information
Hesamoddin Jahanian^1,2, Douglas C. Noll^1,2, Luis Hernandez-Garcia^1,2
1Functional MRI Laboratory, University of Michigan, Ann Arbor, MI, United States; ²Department of Biomedical Engineering, University of Michigan, Ann Arbor, MI, United States

The recent introduction of pseudo-continuous inversion pulses (pCASL) has the potential to greatly facilitate the use of continuous Arterial Spin Labeling (ASL). However, field inhomogeneities, can compromise the tagging efficiency of pCASL, which causes loss in SNR and severe quantification error. We propose a method to restore the loss in labeling efficiency by correcting the phase of the RF pulses in combination with a z-shimming scheme. This will provide more robust perfusion measurements than the conventional pseudo-continuous technique. The method is demonstrated using numerical simulation and In-vivo data.

Robust Prescan for Pseudo-Continuous Arterial Spin Labeling at 7T: Estimation and Correction for Off-Resonance Effects - not available
Wen-Ming Luh¹, S Lalith Talagala², Peter A. Bandettini^1
1FMRIF, NIMH, National Institutes of Health, Bethesda, MD, United States; ²NMRF, NINDS, National Institutes of Health, Bethesda, MD, United States

Pseudo-continuous arterial spin labeling can provide optimal SNR efficiency with sufficient long tag at high fields such as 7T but is very sensitive to off-resonance fields at tagging location as often observed at 7T. Here we demonstrate a robust approach using pair-wise modulation of tagging frequency offset with high SNR images from large voxels and short post labeling delay to derive a necessary ‘prescan’ procedure for estimating and correcting off-resonance effects in 1 minute.

Partial Volume Correction for Perfusion Estimation from Multi-TI Arterial Spin Labelling
Michael A. Chappell^1,2, Adrian R. Groves¹, Bradley J. MacIntosh^1,3, Manus J. Donahue¹, Peter Jezzard¹, Mark W. Woolrich^1
1FMRIB Centre, University of Oxford, Oxford, United Kingdom; ²Institute of Biomedical Engineering, University of Oxford, Oxford, United Kingdom; ³Imaging Research, Sunnybrook Research Institute, Toronto, Canada

The partial voluming of gray matter (GM), white matter (WM) and CSF in ASL leads to underestimates of GM CBF. Here a correction strategy is proposed for multi-TI ASL as part of the kinetic curve model fitting analysis. The method exploits the differences in kinetics between GM and WM and also employs constraints based on partial volume estimates of the tissue types. The proposed method is shown to provide GM CBF estimates corrected for partial voluming while preserving details within the GM CBF image.

Voxel Based Perfusion Variability in ASL
Sanna Gevers¹, Matthias J.P. van Osch², Jeroen Hendrikse³, Reinoud P. Bokkers³, Dennis Kies², Wouter M. Teeuwisse², Charles B.L.M. Majoie¹, Aart J. Nederveen^4
1Radiology, Academic Medical Center Amsterdam, Amsterdam, Netherlands; ²Radiology, Leiden University Medical Center, Netherlands; ³Radiology, University Medical Center Utrecht, Netherlands; ⁴Radiology, Academic Medical Center Amsterdam, Netherlands

Thus far, ASL variability studies have mainly focussed on intrasession and intracenter and multicenter variability of global perfusion and of perfusion in the flow territories of major brain feeding arteries. The purpose of this study was to analyze variability patterns over different brain regions performing a voxel based analysis of variance within and between imaging sessions. The results of our study show that pseudo-continuous ASL with background suppression is least variable over different brain regions whereas other ASL techniques show more variability mainly in vascular regions. Most striking per voxel variances were found in the posterior circulation for pulsed ASL and in the frontal region for continuous ASL.

Superselective Arterial Spin Labeling Applied for Flow Territory Mapping in Selected Clinical Cases - Advantages Over Existing Selective ASL Methods
Michael Helle¹, Matthias van Osch², David Gordon Norris³, Susanne Rüfer¹, Karsten Alfke¹, Olav Jansen^1
1Institute of Neuroradiology, UK-SH, Kiel, Germany; ²C.J. Gorter Center for High Field MRI, Department of Radiology, Leiden University Medical Center, Leiden, Netherlands; ³Donders Institute for Brain, Cognition and Behaviour, Nijmegen, Netherlands

The ability to visualize perfusion territories in the brain is important for many clinical applications but the selectivity of existing methods is restricted to larger vessels. Superselective arterial spin labeling (ASL) is a recently developed technique that overcomes these limitations and permits labeling of small vessels even distal to the Circle of Willis. In this study, superselective ASL is applied for regional perfusion measurements in selected clinical cases (extra-intracranial bypass, arterio-venous malformation and steno-occlusive disease) showing advantages over conventional selective ASL methods and demonstrating benefits in diagnosis, risk analysis and treatment monitoring when added to current MR-protocols.