SRF reported on two AI-CVM studies under the headline “Berner Forschung: KI hilft, Herzkrankheiten genauer vorherzusagen”, highlighting work from the Department of Cardiology, Inselspital, University of Bern, led by Christoph Gräni. The first study, published in The Lancet Digital Health, improved long-term prognostic accuracy in patients with severe aortic stenosis by 10 to 15% compared with standard risk models, using data from 2,985 patients in Bern and external testing in more than 1,000 patients in Japan. The second, published in JAMA Cardiology and conducted with the University of Vienna and several Swiss centres, improved prediction of death and heart failure hospitalisation in transthyretin cardiac amyloidosis by up to 20% over current international staging systems in a cohort of 850 patients. Both models draw on data already collected in routine care, including laboratory values, clinical records and medical imaging, to detect patterns that support individualised risk assessment alongside, and not instead of, clinical judgement.

Link to article: https://www.srf.ch/news/dialog/berner-forschung-ki-hilft-herzkrankheiten-genauer-vorherzusagen