Project

MitraL-AI

Machine learning to support shared decision-making in mitral valve repair

• Healthcare

Challenge

Mitral transcatheter edge-to-edge repair (M-TEER) is a minimally invasive treatment for patients with primary or secondary mitral regurgitation. Patient selection and shared decision-making require a reliable estimate of the long-term prognosis after the procedure.

Approach

Using data collected in clinical centres in Switzerland, Italy and the United States, we developed a machine learning score that predicts all-cause mortality at 12, 24, 36 and 48 months after the procedure, and used explainability methods to study how the importance of each clinical variable changes over the follow-up period.

Impact

A transparent risk score that clinicians and patients can use together when deciding on the treatment.

Publications

• AI-Powered 48-Month Mortality Prediction to Guide Shared Decision-Making in Mitral TEER. The MitraL-AI score: a Swiss, Italian and US collaboration

  L. Biasco, D. Malpetti, L. Azzimonti, C. Berchtold, L. Stanberry, S. Bergstedt, K. Feldewerd, C. Thomas, C. Klersy, A.P. Rubbio, G. Benfari, G. Masiero, P. Biaggi, R. Corti, M. Curti, O. Gaemperli, R. Jeger, F. Maisano, O. Mueller, B. Naegeli, S. Noble, F. Praz, S. Toggweiler, Y. Cristoforetti, G. Musumeci, F. Bedogni, G. Tarantini, L. Gambardella, M. Enriquez-Sarano, G. Pedrazzini

  Manuscript, 2026 Under review