Contributions
Abstract: P800
Type: Poster Sessions
Abstract Category: Pathology and pathogenesis of MS - MRI and PET
Introduction: Widespread cerebral hypoperfusion in Multiple Sclerosis (MS) may contribute to focal lesion formation and diffuse axonal degeneration. Arterial spin labelling (ASL) is a non-invasive MRI method for the assessment of brain perfusion without the need for gadolinium. ASL benefits from higher field strength (7T) due to increased signal-to-noise ratio (SNR) and longer T1 blood and tissue relaxation times. Compared to 2D-EPI ASL, 3D-EPI ASL provides increased spatial coverage, increased SNR, and lower specific absorption rate, with each slice acquired at a constant post-label delay (PLD).
Objectives: To employ 7T 3D-EPI pulsed ASL data in healthy volunteers (HV) and MS patients to determine data quality and feasibility for clinical studies.
Aims: To provide repeatable and robust ASL data for longitudinal clinical studies.
Methods:
Data Acquisition: 12 HVs (mean age 41±8y) and 17 MS patients (mean age 39±7y) were scanned a median of 23 days apart on a 7T Phillips Achieva scanner. Flow alternating inversion recovery (FAIR) ASL data were acquired using a single-shot 3D-EPI readout (spatial resolution=2×2×3mm3, 18 slices, TE/TR=15/42ms, flip angle=14°, SENSE=2 (FH/AP), selective/non-selective width=50/300mm). ASL data were acquired at 5 PLDs, along with an M0 and T1 map using the same imaging parameters for perfusion quantification.
Data Analysis: The T1 map was segmented (FSL FAST) to obtain a grey matter (GM) mask. ASL data for each PLD were motion corrected (FSL MCFLIRT) and co-registered. Perfusion-weighted images were created by subtracting control-label image pairs. Perfusion maps in ml/100g/min were calculated using ASL data from all PLDs and calibrated using the M0 image. The mean GM perfusion and the coefficient of variation (CoV) between scan sessions were calculated.
Results: Data from 1 HV were discarded due to severe head motion. 1 MS patient was deemed an outlier (perfusion difference fell outside the 95% confidence interval). Significant correlation was observed between the GM perfusion values across visits; r=0.74 in HV (CoV=8.4±5.7%) and 0.81 in MS (CoV=8.3±6.3%).
Conclusions: GM perfusion can be quantified reproducibly using a 3D-EPI FAIR ASL at 7T in both HV and MS groups. The observed CoV for both cohorts falls well below 20%, the variability reported in between-session GM perfusion using a FAIR ASL acquisition at the lower field strength of 3T. This protocol is appropriate for longitudinal studies of perfusion changes in MS.
Disclosure: Yasser Falah: nothing to disclose. Richard J Dury: nothing to disclose. Penny A Gowland: nothing to disclose. Nikos Evangelou: nothing to disclose. Susan T Francis: nothing to disclose. Molly G Bright: nothing to disclose.
This work was supported by the Medical Research Council [grant number CiC20160040] and the Anne McLaren Fellowship programme of the University of Nottingham.
Abstract: P800
Type: Poster Sessions
Abstract Category: Pathology and pathogenesis of MS - MRI and PET
Introduction: Widespread cerebral hypoperfusion in Multiple Sclerosis (MS) may contribute to focal lesion formation and diffuse axonal degeneration. Arterial spin labelling (ASL) is a non-invasive MRI method for the assessment of brain perfusion without the need for gadolinium. ASL benefits from higher field strength (7T) due to increased signal-to-noise ratio (SNR) and longer T1 blood and tissue relaxation times. Compared to 2D-EPI ASL, 3D-EPI ASL provides increased spatial coverage, increased SNR, and lower specific absorption rate, with each slice acquired at a constant post-label delay (PLD).
Objectives: To employ 7T 3D-EPI pulsed ASL data in healthy volunteers (HV) and MS patients to determine data quality and feasibility for clinical studies.
Aims: To provide repeatable and robust ASL data for longitudinal clinical studies.
Methods:
Data Acquisition: 12 HVs (mean age 41±8y) and 17 MS patients (mean age 39±7y) were scanned a median of 23 days apart on a 7T Phillips Achieva scanner. Flow alternating inversion recovery (FAIR) ASL data were acquired using a single-shot 3D-EPI readout (spatial resolution=2×2×3mm3, 18 slices, TE/TR=15/42ms, flip angle=14°, SENSE=2 (FH/AP), selective/non-selective width=50/300mm). ASL data were acquired at 5 PLDs, along with an M0 and T1 map using the same imaging parameters for perfusion quantification.
Data Analysis: The T1 map was segmented (FSL FAST) to obtain a grey matter (GM) mask. ASL data for each PLD were motion corrected (FSL MCFLIRT) and co-registered. Perfusion-weighted images were created by subtracting control-label image pairs. Perfusion maps in ml/100g/min were calculated using ASL data from all PLDs and calibrated using the M0 image. The mean GM perfusion and the coefficient of variation (CoV) between scan sessions were calculated.
Results: Data from 1 HV were discarded due to severe head motion. 1 MS patient was deemed an outlier (perfusion difference fell outside the 95% confidence interval). Significant correlation was observed between the GM perfusion values across visits; r=0.74 in HV (CoV=8.4±5.7%) and 0.81 in MS (CoV=8.3±6.3%).
Conclusions: GM perfusion can be quantified reproducibly using a 3D-EPI FAIR ASL at 7T in both HV and MS groups. The observed CoV for both cohorts falls well below 20%, the variability reported in between-session GM perfusion using a FAIR ASL acquisition at the lower field strength of 3T. This protocol is appropriate for longitudinal studies of perfusion changes in MS.
Disclosure: Yasser Falah: nothing to disclose. Richard J Dury: nothing to disclose. Penny A Gowland: nothing to disclose. Nikos Evangelou: nothing to disclose. Susan T Francis: nothing to disclose. Molly G Bright: nothing to disclose.
This work was supported by the Medical Research Council [grant number CiC20160040] and the Anne McLaren Fellowship programme of the University of Nottingham.