Skip to main content
  Sign In   Register

Full Program »

Muscular Mri Pattern Recognition For Muscular Dystrophies: The Era of Artificial Intelligence Beyond A Systematic Review.

Ten years ago, computer researchers thought that getting a computer or a mobile application to tell the difference between two different flowers will be impossible! However, with the significant advance in the state of artificial intelligence (AI) this can be done these days with more than 99% accuracy. Interestingly, AI research within medicine is rapidly growing field, which could be applied for better characterization, diagnosis and prognosis of neuromuscular diseases.

Genetic neuromuscular diseases (gNMD) are a large heterogeneous group of disorders. They comprise more than three hundred genes. Although with the advances on diagnostic modalities, about 40 % of patients with gNMD do not have a diagnosis. Muscle MRI has been proven as a useful tool to orientate the genetic testing by looking at the muscle involvement severity pattern. However, the level of knowledge required to do the proper analysis and eventually differential diagnosis, has limited its use to centers with high degree of expertise.

We believed that the combination of AI- machine learning or image recognition will facilitate the interpretation and comprehension of muscle imagining in hereditary myopathies. To do so, we used an open database containing muscle MRIs scores from 950 individuals and applied AI to use it as a predictor tool. We already have developed an app that can achieve a predictive pattern of more than 90% in a limited number of muscular disorders. This has provided us with the proof of principle that AI can be used in an innovative way in MRI interpretation.

Issa Alawneh
The Hospital for Sick Children
Canada

Hernan Gonorazky
The Hospital for Sick Children
Canada

Sameer Alawna
American University of Sharjah
United Arab Emirates

 


®2002-2021 ICNApedia