Hi folks, I've run into a small problem with HCP script-based processing of non-HCP data that I think is just something that' not yet come up for y'all (i.e., there probably hasn't yet been a need to program into the processing scripts).
As context -- I have fMRI data from a few projects that our Skyra saved only "normalized" (i.e., bias-corrected) data. So I figured I could skip the BiasCorrection step during fMRIVolume processing simply by setting its flag in the GenericfMRIVolumeProcessingPipelineBatch.sh shell to "NONE". This seemed to work fine for one project where I was using SEBASED processing for both T1/T2 and fMRIVolume steps. But for another project it seems advantageous to use gradient echo fieldmap-based distortion correction for the FreeSurfer steps, but spin echo DC on the fMRI data (for minor, technical reasons I won't bore you with but involve circumventing a data acquisition glitch we discovered on our Skyra 6 months into data collection). Using this combination, the HCPs scripts happily get all the way to OneStepResampling before it crashes and says it can't find the BiasField.nii.gz image. It's on Line 135 of the OneStepResampling.sh code that attempts to change the resolution of a non-existent BiasField input image. Can y'all suggest a fix? For instance, should I just put a switch in the OneStepResampling.sh code that skips the code that attempts to change the resolution of a non-existent BiasField input image (line 135 where it's crashing for me)? Or, should I insert code to produce an "empty" BiasField.nii.gz, just so these applywarp and fslmaths steps don't crash? (i.e., not sure where those files are needed later). Or, something else...? Perhaps in an earlier step to produce the relevant BiasField.nii.gz before it gets to OneStepResampling? I'm less certain here what's needed where... I'm assuming there SHOULD be a BiasField.nii.gz image of some sort, empty or not... But I'm not certain where in the HCP scripts it should be produced, nor where subsequent steps expect to have in that image whenever they might call upon it. So I don't want to blindly apply some sort of fix of my own that ends up screwing up the final processed product. I figured y'all could advise to the best way to proceed. Thanks, Mike This e-mail message, including any attachments, is for the sole use of the intended recipient(s) and may contain confidential and privileged information. Any unauthorized review, use, disclosure, or distribution is prohibited. If you are not the intended recipient, or an employee or agent responsible for delivering the message to the intended recipient, please contact the sender by reply e-mail and destroy all copies of the original message, including any attachments. _______________________________________________ HCP-Users mailing list HCP-Users@humanconnectome.org http://lists.humanconnectome.org/mailman/listinfo/hcp-users
