Prerequisites for Custom Model Training

Before you can train a custom model, ensure you have the necessary anatomical masks, tract-density maps. This data should have been prepared according to the Prepare Your Data section. Different from the Predict Mode <prerequisites-predict>`, You will also need the the training labels.

Recommended Data Structure:

A logical and consistent organisation of your files is crucial for a smooth training process. Here’s a suggested structure:

subject001/
├── tracts
│   ├── left
│       ├── seeds_to_target1.nii.gz
│       ├── seeds_to_target2.nii.gz
│   ├── right
│       ├── seeds_to_target1.nii.gz
│       ├── seeds_to_target2.nii.gz
├── roi
│   ├── left
│       ├── tha.nii.gz
│       ├── atlas.nii.gz
│   ├── right
│       ├── tha.nii.gz
│       ├── atlas.nii.gz
├── training-labels
│   ├── left
│       ├── labels.nii.gz
│   ├── right
│       ├── labels.nii.gz
└── otherfiles
subject002/
subject003/

Details:

• tracts directory contains tract-density maps, categorised by hemisphere.

• roi directory stores anatomical masks, also sorted by hemisphere.

• training-labels directory holds the training labels for each hemisphere.

By following this structure, you’ll ensure that the training process is straightforward and efficient.