Skip to content

v0.7.0 Core working pipelines for SLURM
Compare
Choose a tag to compare
@chidiugonna chidiugonna released this 04 Dec 03:59
· 84 commits to main since this release
v0.7.0
702bc48

Introduction

This release provides a reproducible framework for core working pipelines and provides instructions for deploying these pipelines on other SLURM-based high peformance computing (HPC) environments using singularity containers.

Features added

  • Pipelines are arranged sequentially based on dependencies to ensure that jobs are submitted using sbatch using the right dependency parameters
  • Pipelines can also be run locally on a single node that supports singularity however the multiprocessing approach is highly inefficient. This should be used for testing purposes only.
  • Multiple subject sessions now handled by flattening the parallel processing across subject-sessions
  • Issues with FLAIR-based pial reconstruction in Freesurfer version 7.3.2 mitigated in this release.
  • In the Basil pipeline, PLD for PASL acquisitions not correctly handled. The PLD is essentially the TIS for PASL but for PCASL the TIS is calculated as PLD + label duration
  • Group processing pipelines now handled. Currently the collatecsvgroup_panpipeline is the only tested group pipeline which collates neuroimaging measures stored at the single subject level into 1 group csv file.
  • Better logging has been provided using Python's logging module.

Pipelines supported

  • aslprep as aslprep_panpipeline
  • basil as basil_panpipeline
  • freesurfer as freesurfer_panpipeline
  • qsiprep as qsiprep_panpipeline
  • fmriprep as fmriprep_panpipeline
  • amico-noddi as noddi_panpipeline. This pipeline relies on qsiprep_panpipeline
  • DWI tensor metrics as tensor_panpipeline. This pipeline relies on qsiprep_panpipeline and uses MRTrix to calculate tensors like FA, ADC. AD and RD from the preprocessed DWI
  • textmeasures_panpipeline obtains freesurfer structural measures and stores them in a csv file
  • volmeasures_panpipeline obtains neuroimaging measures stored in an image file and stores them in a csv file
  • collatecsv_panpipeline combines all csvs for a single subject and session into 1 csv file.
  • collatecsvgroup_panpipeline combines all colated csvs across the study for all subjects and sessions into 1 single group csv file.
Assets 2
Loading