AI can facilitate treatment of heart attacks 🫀Cardiovascular diseases are the leading cause of death globally. For their timely detection, it’s important to identify fibrosis – scar tissue caused by heart attacks or infectious diseases. Today, clinicians manually process MRI images to measure fibrosis volume, which can be time-consuming.
Researchers from ITMO University and Almazov National Medical Research Centre used deep learning to develop a way to quickly detect cardiac fibrosis. While it takes a clinician about one to two hours to process an image, the model can do it in a few minutes by breaking down the problem into stages:
✅ locate the cardiac muscle
✅ detect fibrosis within each of the 17 traditional heart segments
✅ evaluate its amount in each of them
Currently, the algorithm’s output matches the opinion of two experts in 86% and 77% of cases respectively, which means that the model has human-level performance.
Guest post by ITMO University