Part 2 - Automated Processing

Congratulations! If you’ve completed the previous part successfully you will have a de-identified copy of the IBASH images within your own project on XNAT. The next step is to perform some additional processing on it. Ultimately we want to perform volumetric and cortical thickness analysis on the structural T1 scans. Some of you may have done this previously by downloading data to your local machine and using FreeSurfer. We’re going to avoid creating additional copies of our de-identified data and going through the frustrations of each individual trying to configure these tools locally. Instead we’ll be using some additional features of XNAT, as well as Dockerised versions of common neuroimaging tools to run all our analysis from the XNAT!

In 2-1 Processing with XNAT you’ll be introduced to how XNAT can facilitate running processing pipelines and analyses on data held within its database using the Container Service.

Then in 2-2 DICOM to Nifti and 2-3 FastSurfer you’ll run through manually launching processing and analysis containers from the XNAT interface on imaging objects. The results of your processing will be automatically stored back into XNAT.

After this in 2-4 Defacing you’ll manually launch a defacing algorithm on your T1 Nifti data to further reduce the risk of subject re-identification. To see the impact of defacing on downstream processing, you’ll rerun FastSurfer on this data in order to compare the results to your original output.