 |
 |
Values in grey mater of 100 are not impossible, but I agree that 60-70 in white matter seems rather high. The most likely issues are the M0 value, as you have identified, the default method doesn’t always do a good job of pulling out the pure CSF in the ventricles; or something to do with the setup and the model. Are you calculating the inflow times TIs as the sum of the PLD and the bolus duration? As it stands asl_gui only knows about TIs and not PLDs. Michael On 3 Jun 2015, at 13:00, Julian Schröder <[log in to unmask]> wrote: > Dear ASL experts, > > we are measuring multiple TI pcASL on a Siemens 3T Scanner with currently 10 PLDs (0.3,0.6,0.9 ... 3 s) and use asl_gui to create the perfusion maps. We also aquire M0 calibration data with a TR of 5s (5 volumes). The quantification of absolute perfusion values seems to be inaccurate though. There are values of over 100 ml per 100 g per minute for cortical areas and over 60-70 for white matter, which must certainly be too high. Additionally there are relevant variations in CBF values when measuring the same subject multiple times. > > I have noticed, that the reference tissue CSF mask created automatically by asl_gui, sometimes doesn't fit the respective voxels in the M0 images. Creating a mask manually doesn't improve the results though. > > I hope you can advise me on how to fix these issues. > > with kind regards, Julian Schroeder > > J. Schroeder, M.D. > Department of Neurology > University Medical Center Hamburg-Eppendorf