Intro to SAGIT Concepts

SAGIT workflow organizes around studies. Each SAGIT project is a single study.

The Manifest

Each study should have a manifest that describes where to find imaging data, and what are the conventions.

See Configurations Section for detailed information.

File organization

SAGIT needs to know where to find original data for a study. The minimal data we need for group tractography are:

T1 averaged template
T1 anatomical scans for each subject
Diffusion weighted images for each subject

The default and recommended project folder looks like this:

└── project
    ├── orig
    ├── preprocessed
    ├── rois
    ├── processed
    │   ├── images
    │   └── ...
    └── tractography
        ├── subj1
        ├── subj2
        └── ...

orig: The source folder where original nifti images are found. This include the group template, and the ROI images in template space.
preprocessed: Images created during preprocessing steps are put here, this include the skull stripped images.
rois: This is where the individual T1 space ROI images are created and stored.
processed: This is where all the final template space results are stored. Include the tractography conjunction images.
- processed/images: This is where the visual reports are stored
tractograph: Where the individual tractography are stored. Also include all the individual DWI space files.

Supporting data sources

It's essential for SAGIT to be able to find DWI and Freesurfer segmentation files. The source paths can be configured in the manifest.

SAGIT expect that naming of file and folders match subject names.

If I have a subject.csv file with content:

subject, group
C1, 0
C2, 0
P1, 1
P2, 1

T1 images

T1 images in individual space are generally present in the orig folder. They should be named

C1.nii.gz
C2.nii.gz
...

and so on.

T1 to template transformation sources

SAGIT uses ANTs to register T1 and DWI images. It can autogenerate the T1 to template transforms, but the original T1 images must also be named C1.nii.gz, C2.nii.gz and so on.

DWI sources

DWI images should contain a folder structure such that, Then it should have the folder format:

├── dwi
|   └── C1
└── _C2
    ├── P1
    └── P2

Freesurfer sources

Given the subject.csv file above, the freesurfer should have a folder structure such like this:

├── fsSubjects
|   └── C1
|       ├── mri
|       └── ...
└── _C2
    ├── P1
    └── P2

Basic Concepts