Also if you are going from 3T to 1.5T you should expect some SNR drop. Peace, Matt. _____ From: FSL - FMRIB's Software Library [mailto:[log in to unmask]] On Behalf Of Markus Gschwind Sent: Sunday, April 05, 2009 4:46 PM To: [log in to unmask] Subject: Re: [FSL] What is the Problem with that DTI data? Hi Martin! Your PACS is giving you a strange format! Do you think this should be verified? Nevertheless, in this branch of problem solving I would make sure that the image dataset does not include the trace-weighted images, which Philips calculates if their diffusion encoding scheme is used. This image volume is usually appended at the end of dataset I you can easily identify it, because it looks like diffusion weighted but lacks any WM orientation dependent contrast. I got 1860 DICOM files. After conversion, they end up in 31 volumes of 60 slices. The first volume is the b=0. So far I cannot see a tracer Image. However... I had a detailed look at the dicoms with DICOMWORKS: I was wrong: The acquisition was not on an Achieva, but on a Philips Intera 1.5 T. This could explain the difference in intensity, could it? And I attributed the screenshot wrongly: The scaled image 1:120 is the Philips 1.5T and the scaled image 1:900 is the Trio. Thanks for the bvecs link, I will try it tomorrow. All the best, Markus 2009/4/5 Martin Kavec <[log in to unmask]> Hi Markus, On Sunday 05 April 2009 22:11:56 Markus Gschwind wrote: > Hi Matt and Martin! > > Thanks for help! > > @Martin: > I also have tried dcm2nii. But it does not give me the bvecs and bvals. > They are .IMA files and come from PACS archive. Your PACS is giving you a strange format! Nevertheless, in this branch of problem solving I would make sure that the image dataset does not include the trace-weighted images, which Philips calculates if their diffusion encoding scheme is used. This image volume is usually appended at the end of dataset I you can easily identify it, because it looks like diffusion weighted but lacks any WM orientation dependent contrast. You can get the image slice rotation corrected bvecs by going on the http://godzilla.kennedykrieger.org/~jfarrell/OTHERphilips/GUI.html <http://godzilla.kennedykrieger.org/%7Ejfarrell/OTHERphilips/GUI.html> But for this to use you have to obtain all the information required. This may be difficult if your PACS is stripping away some private DICOM tags. Best, Martin > > @Matt > This is the printscreen of both raw data (same patient, scanned in a > distance of 4 days). > > Obviously the Philips data (attachment, on the right) have much higher > intensity when scaled at the same 0:120 in fslview. > > Is this a reconstruction problem? > > Very grateful! > Markus -- Dr. med. Markus Gschwind, M.D. Laboratory for Neurology and Imaging of Cognition Dept of Neurosciences University Medical Center (CMU) 1 Michel-Servet - 1211 GENEVA - CH Tel 0041 (0) 22 379 5324 Fax 0041 (0) 22 379 5402 email: [log in to unmask] http://labnic.unige.ch