[EPIC] Development of complex analysis workflows

[mrusyd]

Hypothesis Statement

• Funnel Entry Date: 14/6/22
• Epic Name: Development of complex workflows for analysing MR images
• Epic Owner: @tclose

Description

For users of AIS/NIF who acquire MR and/or PET images, AIS-supported pipelines are an integrated analysis service that provides state-of-the-art, convenient, efficient, robust and reproducible methods for medical imaging research data. Unlike ad-hoc workflows, which are either executed manually via CLI or implemented using custom bash/python scripts, our solution is more efficient, robust and reproducible. Unlike BIDS apps, which tend to be monolithic, our solution is more transparent, customisable and amenable to efficient implementation in cloud-based environments.

Outcomes

• Attract new users to the AIS/NIF-facilities by enabling them to analyse datasets in a efficient way
• Reduce load on University compute resources/budget by efficiently executing analyses
• Improve the quality of MR image analyses performed in the NIF network by collaborating/sharing best-practice workflows

Leading Indicators

• User uptake across the network
• Number of pipelines executed

Nonfunctional Requirements

• Compute performance
• Customer satisfaction

Features

Preview Give feedback

1. [FEATURE] Canvassing user requirements #16
   analysis-design feature pipelines +3
   
2. [FEATURE] Implementation of Analysis classes in v2 #14
   deploy-framework mapped pipelines +3
3. [FEATURE] Design structural MRI processing analysis #79
   1 of 7
   analysis-design mapped pipelines +3
4. [FEATURE] Design dMRI preprocessing and analysis #32
   analysis-design pipelines to-map +3
5. [FEATURE] Design functional MRI processing analysis #80
   analysis-design pipelines to-map +3
6. [FEATURE] Design QSM processing analysis #81
   analysis-design pipelines to-map +3
7. [FEATURE] Design MR spectroscopy analysis #82
   analysis-design pipelines to-map +3
8. [FEATURE] Advanced Arcana Analysis class features #52
   deploy-framework mapped pipelines +3

[github-actions]

Stories

• [STORY] Port remaining interfaces required for sMRIPrep - Part I #178
• [STORY] Port ANTs interfaces used in sMRIPrep - part II #149
• [STORY] Port remaining interfaces required for sMRIPrep - Part II #196
• [STORY] Port Freesurfer Recon-all Nipype interface to Pydra #143
• [STORY] Port Freesurfer MRIsConvert Nipype interface to Pydra #198
• [STORY] Port Freesurfer ApplyVolTransform Nipype interface to Pydra #199
• [STORY] Debug QSIPrep BIDS app #206
• [STORY] Finish auto-port script for Nipype -> Pydra workflows #210
• [STORY] Add ability to download OpenNeuro datasets in medimages4tests #213
• [STORY] Port remaining interfaces required for sMRIPrep - Part III #216
• [STORY] Override of field type in task interface auto-port script #221
• [STORY] Create pull-request on Pydra repository regarding mandatory XOR inputs #223
• [STORY] Write unittests for ported pydra-ants interfaces #224
• [STORY] segment T1-weighted anatomical MRI images #230
• [STORY] generate population template of anatomical images #231
• [STORY] deepbet implementation in Pydra #232
• [STORY] segment T1-weighted anatomical MRI images (Bugs to fix) #230 #235
• [STORY] add catalogue of parcellation images to T1 processing pipeline #247
• [STORY] Deploy T1-weighted pipeline as XNAT "app" #251
• [STORY] Merge T1 pipeline sections into a single script #256
• [STORY] Split up Arcana into FrameTree, Pydra2App and core Arcana #314
• [STORY] Write bootstrapping pydra2app command to generate YAML #315

[github-actions]

Features

• [FEATURE] Auto-generated methods section with citations #214
• [FEATURE] T1-weighted pipeline for FRONTIER project #238
• [FEATURE] Porting MRtrix3 Connectome to Pydra #263