AI diagnoses cardiac fibrosis in minutes

Scientists at ITMO and the Almazov National Medical Research Center have created an artificial intelligence-based system that can automatically detect cardiac fibrosis based on magnetic resonance imaging (MRI) data. For this purpose, a special algorithm has been developed that divides the organ image into segments, and then calculates the location and amount of scar tissue.

In the interface area: the digital platform will detect malignant kidney tumors

How a medical decision support service can help doctors and patients

Cardiac fibrosis is an overgrowth of scar tissue that develops after a myocardial infarction or infectious diseases. It is most often diagnosed using MRI, but it takes radiologists an average of one to two hours to accurately measure the volume of pathology in one patient.

The developers told Izvestia that AI can now do this in a matter of minutes. The invention should free doctors from the need to do this manually, which will speed up the selection of the most appropriate treatment strategy for diseases.

"In the proposed algorithm, the user only needs to mark a few points on the image of the heart and classify the sections, and tissue segmentation and generation of the 17-segment diagram are fully automated. We are currently working on improving our method and developing a faster, fully automatic algorithm that will be able to analyze images instantly without user intervention," said Walid Al-Haidri, the main contractor of the project, ITMO researcher.

Other scientific groups have also tried to use a neural network to accomplish this task. At the same time, existing solutions are not precise enough or labor-intensive enough and require the presence of a radiologist, so the need to automate the process of analyzing an MRI image remains.

Read more in the exclusive Izvestia article:

The Truth Segment: AI diagnoses cardiac fibrosis in minutes