Evaluating machine learning muscle ultrasound models on external data

This repository contains model weights that were fine-tuned (using uploaded configs) from the original muscle segmentation models described in the paper "Machine learning-driven Heckmatt grading in facioscapulohumeral muscular dystrophy: A novel pathway for musculoskeletal ultrasound analysis" by Marzola et al. DOI: 10.1016/j.clinph.2025.01.016 The framework and idea behind this work can be found in the paper "Evaluating machine learning muscle ultrasound models on external data – is this the way to propagate forward?" by Błaż et al. DOI: 10.1016/j.clinph.2025.2110968 This repository does not contain any muscle ultrasound images.