New AI model to speed up heart scans, save time, boost treatment

Researchers from the Universities of East Anglia (UEA), Sheffield and Leeds have introduced an innovative AI-driven method for analysing heart MRI scans, aimed at enhancing efficiency and patient care.

The new model, designed to assess heart MRI scans in the four-chamber view, utilises artificial intelligence to swiftly determine the size and function of the heart's chambers. According to Pankaj Garg from UEA's Norwich Medical School, this AI model produces results comparable to those of human experts but in a significantly reduced timeframe. While a traditional MRI scan typically takes 45 minutes, this model completes its analysis in a matter of seconds, promising substantial improvements in clinical practice.

The researchers conducted a retrospective observational study involving 814 patients, supplemented by additional data from 101 patients. Unlike previous studies that focused solely on the heart's two main chambers, this new approach examines all four chambers.

The team anticipates that this advancement will lead to faster and more precise diagnoses, ultimately benefiting patients. They emphasised the importance of future studies involving larger and more diverse patient groups across various hospitals and MRI scanner types to validate its accuracy in broader clinical settings.