Machine learning enhances predictions for biliary tract cancer survival.

A novel “biliary tract cancer survival calculator” was developed, utilizing machine learning to provide personalized survival estimates for patients with biliary tract cancers, incorporating data from 62,877 cases. Key factors negatively impacting survival included metastatic disease, age at diagnosis, and lack of surgical treatment. The model demonstrated significant accuracy during internal validation, with a c-index of 0.74, outperforming traditional stage-based methods (c-index of 0.64), thereby enabling real-time prognostic assessments.