Browse by author
Lookup NU author(s): Jose Verdú-DíazORCiD, Carla Bolaño DiazORCiD, Alejandro Gonzalez Chamorro, Sam Fitzsimmons, Dr Goknur Kocak, Dr Kieren HollingsworthORCiD, Professor Chiara Marini Bettolo, Professor Volker StraubORCiD, Professor Giorgio TascaORCiD, Professor Jaume BacarditORCiD, Dr Edith Diaz-Mireles
This work is licensed under a Creative Commons Attribution 4.0 International License (CC BY 4.0).
© 2025 The Author(s). Journal of Cachexia, Sarcopenia and Muscle published by Wiley Periodicals LLC.Background: Neuromuscular diseases (NMDs) are rare disorders characterized by progressive muscle fibre loss, leading to replacement by fibrotic and fatty tissue, muscle weakness and disability. Early diagnosis is critical for therapeutic decisions, care planning and genetic counselling. Muscle magnetic resonance imaging (MRI) has emerged as a valuable diagnostic tool by identifying characteristic patterns of muscle involvement. However, the increasing complexity of these patterns complicates their interpretation, limiting their clinical utility. Additionally, multi-study data aggregation introduces heterogeneity challenges. This study presents a novel multi-study harmonization pipeline for muscle MRI and an AI-driven diagnostic tool to assist clinicians in identifying disease-specific muscle involvement patterns. Methods: We developed a preprocessing pipeline to standardize MRI fat content across datasets, minimizing source bias. An ensemble of XGBoost models was trained to classify patients based on intramuscular fat replacement, age at MRI and sex. The SHapley Additive exPlanations (SHAP) framework was adapted to analyse model predictions and identify disease-specific muscle involvement patterns. To address class imbalance, training and evaluation were conducted using class-balanced metrics. The model's performance was compared against four expert clinicians using 14 previously unseen MRI scans. Results: Using our harmonization approach, we curated a dataset of 2961 MRI samples from genetically confirmed cases of 20 paediatric and adult NMDs. The model achieved a balanced accuracy of 64.8% ± 3.4%, with a weighted top-3 accuracy of 84.7% ± 1.8% and top-5 accuracy of 90.2% ± 2.4%. It also identified key features relevant for differential diagnosis, aiding clinical decision-making. Compared to four expert clinicians, the model obtained the highest top-3 accuracy (75.0% ± 4.8%). The diagnostic tool has been implemented as a free web platform, providing global access to the medical community. Conclusions: The application of AI in muscle MRI for NMD diagnosis remains underexplored due to data scarcity. This study introduces a framework for dataset harmonization, enabling advanced computational techniques. Our findings demonstrate the potential of AI-based approaches to enhance differential diagnosis by identifying disease-specific muscle involvement patterns. The developed tool surpasses expert performance in diagnostic ranking and is accessible to clinicians worldwide via the Myo-Guide online platform.
Author(s): Verdu-Diaz J, Bolano-Diaz C, Gonzalez-Chamorro A, Fitzsimmons S, Warman-Chardon J, Kocak G, Mucida-Alvim D, Smith I, Vissing J, Poulsen N, Luo S, Dominguez-Gonzalez C, Bermejo-Guerrero L, Gomez-Andres D, Sotoca J, Pichiecchio A, Nicolosi S, Monforte M, Brogna C, Mercuri E, Bevilacqua J, Diaz-Jara J, Pizarro-Galleguillos B, Krkoska P, Alonso-Perez J, Olive M, Niks E, Kan H, Lilleker J, Roberts M, Buchignani B, Shin J, Esselin F, LeBars E, Childs A, Malfatti E, Sarkozy A, Perry L, Sudhakar S, Zanoteli E, DiPace F, Matthews E, Attarian S, Bendahan D, Garibaldi M, Fionda L, Alonso-Jimenez A, Carlier R, Okhovat A, Nafissi S, Nalini A, Vengalil S, Hollingsworth K, Marini-Bettolo C, Straub V, Tasca G, Bacardit J, Diaz-Manera J, Myo-Guide Consortium
Publication type: Article
Publication status: Published
Journal: Journal of Cachexia, Sarcopenia and Muscle
Year: 2025
Volume: 16
Issue: 3
Print publication date: 01/06/2025
Online publication date: 24/04/2025
Acceptance date: 25/03/2025
Date deposited: 13/05/2025
ISSN (print): 2190-5991
ISSN (electronic): 2190-6009
Publisher: John Wiley and Sons Inc
URL: https://doi.org/10.1002/jcsm.13815
DOI: 10.1002/jcsm.13815
Data Access Statement: The corresponding author can facilitate a list of the publications with data included in this work. Unpublished data are not eligible for sharing due to a lack of consent from patients. Most of the dysferlinopathy data belong to the Jain COS project, available at www.jain-foundation.org. The code used for this project is published under the myoguide-diagnose python package, available at www.github.com/MYO-Guide/myoguide-diagnose .
PubMed id: 40275674
Altmetrics provided by Altmetric
