Thanks to the use of artificial intelligence, a new diagnostic technique of magnetic resonance for the diagnosis and monitoring of rare neuromuscular diseases arrives in Italy. In a few seconds, up to a maximum of ten, it is possible to obtain accurate information on the properties of pathological tissues.
The experimental use of neural networks for the acquisition of magnetic resonance images is the object of the research project developed at the Mondino Foundation in Pavia by 30-year-old Leonardo Barzaghi and 26-year-old Raffaella Fiamma Cabini, PhD students at the Mondino “BioData Science” Center , coordinated by Professor Silvia Figini. It was they, already graduates in physics from Pavia and Milan, who presented it in world preview at the CompMat Spring Workshop”, the event dedicated to the new frontiers of machine learning and computational mathematics, organized today by the University of Pavia and held in the Foscolo Hall.
Thanks to two scholarships funded by the neurological center of Pavia, the two researchers studied the development of “machine learning” and “deep learning” algorithms for the prediction of quantitative biomarkers of diseases of the musculoskeletal system. “To support the diagnosis and increase the quality of the high-resolution anatomical images obtained through magnetic resonance, techniques have been developed in recent years that allow us to quantify the physical properties of pathological tissues – explains Barzaghi -. The use of the most advanced models of artificial intelligence allows today to accelerate the acquisition times of quantitative information of the pathology such as, for example, those relating to the amount of inflammation, atrophy and the percentage of fat.Thanks to neural networks, images can be obtained in a few seconds, reducing the necessary times with standard methods, quantifiable in hours”. Barzaghi’s research focused on clinical images, while Cabini’s on preclinical ones.
breaking latest news © Copyright ANSA
