Active Bones, Joints & Muscles NIHR-supported project Brain & Nervous System

MyoGuide+: accelerating the diagnosis of patients with neuromuscular diseases applying artificial intelligence to muscle MRIs

In plain English

AI plain-English summary

A free online tool now uses artificial intelligence to predict which of 20 different muscle-wasting diseases a patient has by analysing patterns of fatty replacement in their MRI scans. Diagnosing neuromuscular diseases like muscular dystrophy often takes years, as symptoms overlap and specialist expertise is scarce. Doctors currently rely on visual inspection of muscle MRIs, a slow and subjective process. MyoGuide addresses this gap by applying machine learning to detect disease-specific patterns of fat infiltration in skeletal muscles, offering a rapid, standardised second opinion. The research team is now expanding the tool to cover more diseases and additional muscle groups, including the trunk, upper limbs, and cranial muscles. They are also building an automatic segmentation tool to quantify fat replacement without manual input, and adding an atlas of disease patterns alongside educational videos for clinicians, patients, and families. If successful, MyoGuide could cut diagnostic delays from years to weeks, allowing earlier access to treatments and clinical trials for patients with rare neuromuscular conditions. It also demonstrates how AI can turn routine clinical imaging into a diagnostic aid for diseases that are otherwise difficult to identify.

View original technical description
Myo-Guide represents a cutting-edge tool exploiting the power of machine learning to accurately predict the diagnosis of up to 20 distinct neuromuscular diseases. This is achieved through the analysis of fatty replacement patterns in skeletal muscles, as assessed by muscle MRI. After years of dedicated development, Myo-Guide is now accessible free of charge through the web platform myoguide.org, with the primary objective of assisting healthcare providers in efficiently reaching diagnoses for patients. Our commitment to advancing medical diagnostics propels us to enhance Myo-Guide further. The ongoing efforts include broadening its predictive capabilities to encompass additional diseases expanding the anatomical coverage to include the trunk, upper limbs, and cranial muscles. Additionally, we seek to incorporate information regarding inflammatory changes within muscles, discernible through the utilization of Short Tau Inversion Recovery sequence. A pivotal aspect of this evolution involves the creation of an automatic segmentation tool, to identify and quantify fat replacement of muscles, with the aim to significantly facilitate the diagnostic workflow for clinicians using the system. Furthermore, we aim to refine the web interface housing the machine learning tool. The forthcoming updates encompass a more intuitive and user-friendly interface tailored specifically for clinicians. The introduction of a novel segmentation tool stands as a pivotal enhancement, complemented by the inclusion of an atlas detailing the patterns observable in patients with different diseases. Moreover, we am to incorporate educational resources, including videos and documents, to benefit clinicians, patients, their families, and the general public.

Researchers

Jordi Diaz Manera (Principal Investigator)

Related Research

Grants with similar aims, by meaning.

Implementation of an artificial intelligence module on the online imaging portal MYO-Share for guiding the diagnosis of muscle diseases.
Implementation of an artificial intelligence module on the online imaging portal MYO-Share for guiding the diagnosis of muscle diseases
Implementation of an artificial intelligence module on the online imaging portal MYO-Share for guiding the diagnosis of muscle disease
21BI49 - Implementation of an artificial intelligence module on the online imaging portal MYO-Share for guiding the diagnosis of muscle diseases
Automating muscle MRI analysis with Artificial Intelligence for the diagnosis and study of neuromuscular diseases on a web application

Original classification

Neuromuscular Disease, Rare Diseases & Mitochondrial Dysfunction

Plain English summaries and category classifications on this site are generated by AI and may not perfectly reflect the original research.