Skip to content

workflow.md

Latest commit

 

History

History
55 lines (47 loc) · 1.42 KB

workflow.md

File metadata and controls

55 lines (47 loc) · 1.42 KB
Raw

Workflow

This repository is organized around one public entry point: bin/aumc-convert. The lower-level scripts are intentionally small so that AUMC-specific rules can be added without turning the whole workflow into one large shell file.

source imaging data
       │
       ▼
argument parsing
       │
       ▼
input + subject validation
       │
       ▼
resolve output directory
       │
       ▼
conversion backend
(default: dcm2niix)
       │
       ▼
subject[/session] output
       │
       ▼
summary / downstream QC

Responsibilities

Component Responsibility
bin/aumc-convert CLI, orchestration, subject/session naming
scripts/validate_input.sh Fail early on unsafe or invalid inputs
scripts/convert_subject.sh Converter command construction and execution
scripts/helpers.sh Logging, config loading, path helpers
config/local.env Machine-specific configuration; never committed

Where to add AUMC-specific logic

If the existing AUMC workflow contains sequence mapping, renaming, metadata patching, or post-conversion organization, place each concern in its own script, for example:

scripts/
├── convert_subject.sh
├── map_sequences.sh
├── patch_metadata.sh
└── organize_output.sh

Then call those stages from bin/aumc-convert in order. This keeps the public CLI stable even if the implementation changes.