Machine learning scans MRIs to diagnose specific forms of muscular dystrophy

A new artificial intelligence tool can accurately distinguish specific forms of muscular dystrophy by analyzing MRI scans, according to recent research out of Spain.

Diagnosing this muscle degeneration disease usually relies on genetic testing, researchers wrote in the journal Neurology. Clinical findings typically help guide the examination, including magnetic resonance imaging of where fat tissue has replaced muscle tissue. But narrowing a final conclusion down to a specific form of MD can be quite challenging.

José Verdú-Díaz, with Copenhagen Neuromuscular Center’s Department of Neurology, and colleagues used hundreds of images to create their machine learning tool, which achieved a diagnostic accuracy of 95.7%, beating out a group of expert radiologists.

“This tool can be of great help in the diagnostic process of MDs, providing potential genes to be tested or reinforcing the pathogenic role of a mutation found,” the authors wrote.

In order to create their AI, the researchers collected 976 pelvic and lower limb muscle MRIs, detailing 10 different forms of muscular dystrophy. They then generated 2,000 different models and selected the most accurate one.

When Verdú-Díaz et al. tested this approach on a new set of 20 test MRIs, it achieved a 95.7% accuracy, 92.1% sensitivity and 99.4% specificity.

Although their research is in its early phases, the authors believe they have taken a positive step forward for those suffering from muscular dystrophy.

“This study can be considered as a proof of concept that demonstrates that artificial intelligence can be applied to the field of muscle MRI,” they concluded.