emilljungberg / fw-SamSeg

SamSeg

Sequence Adaptive Multimodal SEGmentation (SAMSEG) is a tool to robustly segment dozens of brain structures from head MRI scans without preprocessing. The characteristic property of SAMSEG is that it accepts multi-contrast MRI data without prior assumptions on the specific type of scanner or pulse sequences used.

Overview

Usage

FAQ

Summary

Cite

license:

url: https://gitlab.com/flywheel-io/flywheel-apps/

cite:
Fast and sequence-adaptive whole-brain segmentation using parametric Bayesian modeling. O. Puonti, J.E. Iglesias, K. Van Leemput. NeuroImage, 143, 235-249, 2016.

Classification

Category: analysis

Gear Level:

• Project
• Subject
• Session
• Acquisition
• Analysis

---

Inputs

• api-key
  • Name: api-key
  • Type: object
  • Optional: true
  • Classification: api-key
  • Description: Flywheel API key.

Config

• debug

  • Name: debug
  • Type: boolean
  • Description: Log debug messages
  • Default: false

• input

  • Base: file
  • Description: input file (usually isotropic reconstruction)
  • Optional: false

Outputs

• output

  • Base: file
  • Description: segmentated file
  • Optional: false

• parcelation

  • Base: file
  • Description: parcelation file
  • Optional: true

• vol

  • Base: file
  • Description: volume estimation file (csv)
  • Optional: true

• QC

  • Base: file
  • Description: QC file (csv)
  • Optional: true

Metadata

No metadata currently created by this gear

Pre-requisites

• Three dimensional structural image

Prerequisite Gear Runs

This gear runs on BIDS-organized data. To have your data BIDS-ified, it is recommended that you run, in the following order:

1. dcm2niix
   • Level: Any
2. file-metadata-importer
   • Level: Any
3. file-classifier
   • Level: Any

Prerequisite

Usage

This section provides a more detailed description of the gear, including not just WHAT it does, but HOW it works in flywheel

Description

This gear is run at either the Subject or the Session level. It downloads the data for that subject/session into the /flwyhweel/v0/work/bids folder and then runs the synthseg pipeline on it.

After the pipeline is run, the output folder is zipped and saved into the analysis container.

File Specifications

This section contains specifications on any input files that the gear may need

Workflow

A picture and description of the workflow

  graph LR;
    A[T1w]:::input --> FW;
    FW[FW] --> FMI;
    FMI((file-metadata-importer)):::gear --> FC;
    FC((file-classifier)):::gear --> D2N;
    D2N((dcm2niix)):::gear --> CB;
    CB((curate-bids)):::gear --> CISO;
    CISO((ciso)):::gear --> SS;
    SS((synthseg)):::gear --> ANA;
    ANA[Analysis]:::container;
    
    classDef container fill:#57d,color:#fff
    classDef input fill:#7a9,color:#fff
    classDef gear fill:#659,color:#fff

Loading

Description of workflow

1. Upload data to container
2. Prepare data by running the following gears:
   1. file metadata importer
   2. file classifier
   3. dcm2niix
   4. MRIQC (optional)
   5. curate bids
3. Select either a subject or a session.
4. Run the ciso gear (Hyperfine triplane aquisitions)
5. Run the synthseg gear
6. Gear places output in Analysis

Use Cases

FAQ

FAQ.md

Contributing

[For more information about how to get started contributing to that gear, checkout CONTRIBUTING.md.]