Spinal Cord Toolbox (SCT) - New release v7.0

joshuacwnewton · April 28, 2025, 2:34pm

Dear neuroimaging community,

We are happy to announce the release of SCT v7.0.

The most notable changes for this major release include:

Feature: Overhauled sct_deepseg syntax, interface, and model names. View Model Gallery and pull request
- Add lesion_ms_axial_t2 model to sct_deepseg for MS lesion/SC segmentation of axial T2 images. View pull request
- Add graymatter model to sct_deepseg for contrast-agnostic segmentation of the gray matter. View pull request
- Update spinalcord contrast-agnostic model to retrain from scratch using nnUNetv2 framework. View pull request
Feature Add new CLI tool (sct_detect_compression) to predict compression probability. View pull request
Feature: Add -lrootlet argument to sct_register_to_template to enable rootlets-informed registration. View pull request
Feature: Add detailed time and memory profiling to all CLI scripts. View pull request
Enhancement: Overhaul the QC report backend to use modern web technologies. View pull request
Enhancement: Update PAM50 to include changes to PAM50_rootlets.nii.gz (ventral rootlets, Th1 level). View pull request
Enhancement: Recompute PAM50 normalized metrics using the newest spinalcord segmentation model. View pull request
Installation: Adopt Miniforge environment manager to address Miniconda licensing concerns. View pull request
Documentation: Update web tutorials to match the changes made for the 2024 SCT Course. View pull request
Testing: Add new GitHub Actions workflow to ensure that older releases install without error. View pull request

For the full list of changes included in this release, click here.

For installation instructions, click here.

If you have any questions or feature requests, please post in the SCT forum.

Happy processing!

The Spinal Cord Toolbox Team