Teresa Matias Correia, a researcher at the Centro de Ciências do Mar do Algarve (CCMAR), together with scientists from Portugal and Spain, is optimizing a new technique that combines magnetic resonance imaging and artificial intelligence to prematurely identify coronary heart disease, the leading cause of death in the world.
“Currently, the most widely used method to detect this disease is coronary angiography, an invasive technique that uses X-rays to view the blood vessels of the heart,” explains Correia. “In addition to being expensive, it is a complex technique that requires specialized personnel and hospitalization. We are looking for non-invasive and non-radiation alternatives, such as cardiac perfusion magnetic resonance imaging; a technique that already exists, but still poses several limitationsâ€, explains the researcher.
CMR captures how blood flows in the heart. However, it has low resolution, does not capture the entire organ, and has room for interpretation, depending on the level of training of the person viewing the images. To overcome these obstacles, Correia is finalizing a new MRI technique that combines mathematical methods and machine learning to speed up the process, in order to allow more accurate diagnoses with little room for discussion about whether a person has coronary heart disease or is developing it.
“We are working with a team from the Instituto Superior Técnico de Lisboa who are programming the scanner to change the way we acquire images. Also with the University of Valladolid, which works on motion correction and heart image processing. In addition, we are collaborating with the National Center for Cardiovascular Research in Madrid for clinical validationâ€, says Correia. “By combining mathematics, physics and artificial intelligence, we are changing the way data are obtained and conclusions about heart health to reduce morbidity and mortalityâ€, concludes the researcher.
The Ask Big Vang section is financed by the La Caixa Foundation.
