v0.9.3 Core working pipelines

Changes in this release

• Add new nilearn-based measure extraction pipeline, workflow and node which enables a slightly more robust way of handling missing values. The roiextract_panpipeline, roiextract_workflow and roi_extract.py should be used instead of the volmeasures_panpipeline, volmeasures_workflow and the roi_mean_single nodes.
• Use bids-like nomenclature for atlases and associated metadata for use in the new roi_extract workflow.
• Use bids-id to identify subjects for array jobs and include new tag IN-XNAT. Together these changes will allow local bids subjects to be processed if they are not downloadable from XNAT and opens path for other ways for downloading data e..g. rsync, sftp, AWS etc...
• Use sdcflow library to directly produce fieldmaps from MEGRE for Basil workflow. This is provided via the script interface using the FMRIPREP container for compatibility with the fmriprep created fieldmaps (sdcflow==1.8.0). However the PAN_CONTAINER can be used as well. (sdcflow == 1.8.1)
• Several changes to the way that atlas indices are generated for new and existing atlases. ANTS transforms also performed as floats and then subsequently converted to int for atlases using FSL.
• PAN250 Deployment folder created for better reproducibility.