Link to Source: Github, Paper, Online Web Application
Summary: Predicts long-term mortality after TAVI procedures using multimodal AI and clinical data analysis
TAVI-ALPURA is an artificial intelligence-based clinical decision support tool that predicts long-term all-cause mortality and cardiovascular death following transcatheter aortic valve implantation (TAVI) procedures. Named to reflect its Swiss and Japanese origins (combining “ALP” from the Swiss Alps and “URA” from Sakura, Japan’s cherry blossom), the system was developed on a Swiss cohort and externally validated in a Japanese cohort, demonstrating superior performance compared to conventional risk scores for long-term outcome prediction. The tool’s multimodal approach integrates diverse clinical data sources to provide comprehensive risk stratification.
Accessible through a user-friendly web application at ai-cvm.com/software/tavi-alpura, TAVI-ALPURA demonstrated strong generalizability across diverse patient populations, making it a valuable tool for clinicians managing TAVI patients. The system is currently under review for publication and is intended strictly for research purposes. By leveraging advanced machine learning techniques on multicenter data, TAVI-ALPURA aims to improve patient selection, counseling, and post-procedure monitoring strategies, ultimately contributing to better outcomes for patients undergoing this increasingly common cardiac intervention.