It's time for a Q&A! Can you explain what steps you’ve taken to adapt the pipelines to work with > the BIDS format? > The pipelines are not modified - BIDS App is calling exactly the same high level functions as the scripts in https://github.com/Washington- University/Pipelines/tree/master/Examples are calling. All of the of work of adapting to BIDS is figuring out which raw files go where and which parameters to set. All of this code is contained in this script: https://github.com/BIDS-Apps/HCPPipelines/blob/master/run.py
> Conversely, have you taken any steps to adapt the existing HCP packages to > work with these containers? > If by "HCP packages" you mean datasets than yes. As mentioned in the readme "To convert DICOMs from your HCP-Style (CMRR) acquisitions to BIDS try using heudiconv <https://github.com/nipy/heudiconv> with this heuristic file <https://github.com/nipy/heudiconv/blob/master/heuristics/cmrr_heuristic.py>.". We are also working on a simple HCP2BIDS tool (I'll posted it here when it's ready). This indeed looks very interesting. Following up on Dan’s question, it > will be important for users to have a transparent way to know if any > changes have been made to the HCP Pipeline scripts. > I wanted to avoid having to change anything in the original Pipelines. Thus I'm only calling existing top level scripts (like PreFreeSurfer.sh) with command line parameters inferred from the input BIDS dataset. Is the part above that “lack of fieldmaps” is handled “robustly” only in > reference to the T1/T2 scans? Because the HCP Pipelines have always > considered fieldmaps to be a mandatory acquisition for the fMRI processing > (given their importance in correcting for distortions). > It applies to both structural as well as functional scans. The word "robustly" is probably not the best in this context. All I mean is that the App will still run (by passing "NONE" to the HCP scripts which as far a I know is a valid option - please correct me if I am wrong). The results will of course be worse if fieldmaps were present. I would appreciate if you could suggest a better wording that would cause less confusion. Let me know if you have any other questions/suggestions/comments. Best, Chris > cheers, > -MH > > -- > Michael Harms, Ph.D. > ----------------------------------------------------------- > Conte Center for the Neuroscience of Mental Disorders > Washington University School of Medicine > Department of Psychiatry, Box 8134 > 660 South Euclid Ave. Tel: 314-747-6173 > St. Louis, MO 63110 Email: [email protected] > > From: <[email protected]> on behalf of Daniel Marcus < > [email protected]> > Date: Monday, September 12, 2016 at 2:27 PM > To: Chris Gorgolewski <[email protected]>, Hcp Users < > [email protected]>, "[email protected]" < > [email protected]> > Subject: Re: [HCP-Users] HCP Pipelines BIDS App > > Hi Chris, > > This is great work and I think will be a serious asset to those looking to > run HCP pipelines on their data. Can you explain what steps you’ve taken > to adapt the pipelines to work with the BIDS format? Conversely, have you > taken any steps to adapt the existing HCP packages to work with these > containers? > > -Dan > > From: <[email protected]> on behalf of Chris > Gorgolewski <[email protected]> > Date: Sunday, September 11, 2016 at 12:03 AM > To: Hcp Users <[email protected]>, " > [email protected]" <[email protected]> > Subject: [HCP-Users] HCP Pipelines BIDS App > > Dear HCP community, > > Many researchers want to run HCP Pipelines on their own datasets, but run > into problems with software dependencies and different organization of data > and metadata. To help with this issue I have created an HCP Pipelines BIDS > App. Like all other BIDS Apps <http://bids-apps.neuroimaging.io/> the HCP > Pipelines one has the following features: > > - It's portable (meaning comes with all of the dependencies with the > correct versions - no need to install FSL or Freesurfer). > - Runs on Windows, Mac OS X and Linux (as well as HPCs or clusters via > Singularity <http://bids-apps.neuroimaging.io/tutorial/>). > - No need to specify any metadata - all you need as an input is a BIDS > <http://bids.neuroimaging.io/>dataset . > - The only software required (across all three platforms) is Docker > <https://www.docker.com/>. > - The App (which includes all dependencies) is versioned and all > historical versions are preserved. This allows you to keep the same > software stack intact during a longitudinal study spread over years. > > The App was designed to parse the input datasets, figure out which scans > are available and run HCP pipelines with optimal parameters. I have tested > it with the HCP example subject, but if you have data you would like to > provide for testing purposes I would be very happy to take advantage of it. > > I hope this will make HCP pipelines accessible to more researchers. Please > let me know what you think! > > More information how to use the HCP Pipeline App can be found here: > https://github.com/BIDS-Apps/HCPPipelines > > Best, > Chris Gorgolewski > > _______________________________________________ > HCP-Users mailing list > [email protected] > http://lists.humanconnectome.org/mailman/listinfo/hcp-users > > > ------------------------------ > > The materials in this message are private and may contain Protected > Healthcare Information or other information of a sensitive nature. If you > are not the intended recipient, be advised that any unauthorized use, > disclosure, copying or the taking of any action in reliance on the contents > of this information is strictly prohibited. If you have received this email > in error, please immediately notify the sender via telephone or return mail. > > _______________________________________________ > HCP-Users mailing list > [email protected] > http://lists.humanconnectome.org/mailman/listinfo/hcp-users > > > ------------------------------ > > The materials in this message are private and may contain Protected > Healthcare Information or other information of a sensitive nature. If you > are not the intended recipient, be advised that any unauthorized use, > disclosure, copying or the taking of any action in reliance on the contents > of this information is strictly prohibited. If you have received this email > in error, please immediately notify the sender via telephone or return mail. > _______________________________________________ HCP-Users mailing list [email protected] http://lists.humanconnectome.org/mailman/listinfo/hcp-users
