SummaryThis project aims to improve how muscle health is assessed in people with neurological conditions. These conditions can cause significant changes in muscle structure and composition, which affect strength and mobility. MRI is one of the best tools for measuring these changes, but traditional analysis methods often lack ability to identify coherent patterns. We will use radiomics—advanced techniques that extract detailed information from MRI images—together with explainable artificial intelligence (AI) methods such as SHAP. This combination will help identify which muscle features are most important and provide clear, transparent explanations for how the analysis works. The project builds on a previously validated radiomics framework and will test its use on both a large population dataset and a smaller clinical dataset. The expected outcome is a more interpretable and reliable way to detect muscle changes linked to neurological disorders.

Keywords MRI Radiomics explainable-AI muscle-quality

InstitutionsRoyal Institute of Technology