You can deploy this model and start earning money today!

yeshekway/
lower-limb-muscle-segmentation

Name: lower-limb-muscle-segmentation
Availability: InStock

creativeml-openrail-m

Views

No views yet

Lower-Limb Muscle and Adipose Tissue Segmentation Cascade

This repository contains 3D deep learning models trained on two-point Dixon MRI sequences for automated segmentation of lower-limb muscles and subcutaneous adipose tissue (SAT). The models form a cascaded pipeline for precise quantification of muscle volumes and fat distribution.

Models

lower-limb-muscle-and-sat-segmentation.pth
Segments total skeletal muscle area and subcutaneous adipose tissue in the lower leg, calf and thigh volumes.
calf-muscle-segmentation.pth
Segments nine individual calf muscles.
thigh-muscle-segmentation.pth
Segments thirteen individual thigh muscles.

Code

Full preprocessing, training, and inference scripts are available on GitHub:
GitHub Repository

Refer to the Final app folder in the GitHub repo to run the full cascade with these pretrained models.

Paper

For full details of the methodology, datasets, and validation, please see our paper: Springer Article

Dataset

The dataset used to train these models is publicly available at: Figshare DOI

License

CreativeML OpenRAIL-M

yeshekway/lower-limb-muscle-segmentation