Guides

User Guide

• Installation and Setup Guide
  • 1. Install from the GitLab package registry (recommended)
  • 2. Source / Editable Install (development)
  • 3. Docker Container (CPU and GPU)
  • 4. HPC via Apptainer
  • Configuration
  • Running Tests
  • Building Documentation
  • Development Setup
  • Troubleshooting
• HPC with Apptainer
  • Why containers on HPC
  • Prerequisites
  • Storage layout
  • A shared env.sh
  • Installing the tool containers
  • Conda environment for the pipeline
  • FreeSurfer license
  • Verification
  • Reference: idempotent setup script
• Workflow Usage Guide
  • Tractography backends
  • End-to-end meta-pipelines
  • Preprocessing, registration, segmentation
  • Cohort, atlas, similarity
  • QC and examples
  • Common usage patterns
  • Output expectations
  • Choosing a workflow
  • See also
• CLI Reference
  • Top-level group
  • thesis run
  • thesis list-workflows
  • thesis list-configs
  • thesis show-config
  • thesis info
  • thesis stats collect
  • Exit codes
  • CLI ↔ config mapping
• Configuration Guide
  • Merge order
  • Path syntax
  • Example structure
  • Loading configuration from Python
  • Inspecting the merged config
  • Patient-specific overrides
  • CLI overrides
  • Which sections does my workflow use?
  • Per-section reference
  • Best practices
• Troubleshooting
  • First Checks
  • Missing External Tools
  • Missing Input Files
  • Config Problems
  • Nipype Runtime Issues
  • MRtrix3 Gotchas
  • Documentation and Developer Checks
  • Workflow Preflight Errors
  • Developer / Extension Errors
  • Still Stuck?
  • Documentation and Developer Checks

Extending the Framework

• Project Architecture Overview
  • Package Structure
  • Design Patterns
  • Core Module Responsibilities
  • Exception Hierarchy
  • Data Flow
  • Adding a New Workflow
  • Key Principles
• Nipype Integration Guide
  • Overview
  • Minimal Pattern
  • Integration Layers
  • Current Architecture
  • Building Workflow Modules
  • Execution Model
  • Workflow I/O contracts
  • Nested Workflows
  • Custom Interfaces and Adapters
  • Multi-Source ROI Handling in HCP
  • Configuration Notes
  • Troubleshooting
  • Related Documentation
• Nipype Architecture Diagrams
  • 1. End-to-End Control Flow
  • 2. Runtime Layers
  • 3. HCP Workflow Graph Shape
  • 4. HCP + SynthSeg Meta-Workflow
  • 5. Progress Reporting
  • 6. Directory Layout
  • 7. Config to Execution Flow
  • 8. Key Design Choice
• Building a workflow
  • TL;DR
  • The four decorators
  • Path declarations
  • A complete, annotated example
  • ProcessingContext: the state carrier
  • inputnode / outputnode contracts (for composability)
  • Logging rule (and the Function-node exception)
  • Owning a config section
  • Registration and discovery
  • Common gotchas
  • Runnable examples
  • Where to look next
• Adapting an arbitrary Nipype script to the framework
  • The single biggest shift
  • Step 1 — Wrap the script body in a single @workflow function
  • Step 2 — Replace hardcoded input paths with @requires / PatientFile
  • Step 3 — Replace DataSink / manual output dirs with @produces / OutputDir
  • Step 4 — Thread subject identity through ProcessingContext, not globals
  • Step 5 — Remove sys.path / importlib hacks and sibling-module imports
  • Step 6 — Declare a Pydantic schema for the workflow’s YAML block
  • Step 7 — Fix logging (print inside Function nodes)
  • Putting it together — the assembled file
  • Step 8 — Register and run it
  • Verify it runs
  • Troubleshooting
  • See also
• Best Practices for This Framework
  • Code Organization
  • Type Hints and Documentation
  • Logging
  • Testing
  • Error Handling
  • Configuration Best Practices
  • Code Quality
  • Performance Considerations
  • Documentation
  • Git Workflow
• Contributing to Thesis Imaging Pipeline
  • Prerequisites
  • Getting Started
  • Development Workflow
  • Pull Request Process
  • Adding a New Workflow
  • Running Existing Tests
  • Documentation
  • Reporting Issues
  • Getting Help
  • Code of Conduct