MED-ADVANCE
Personalized Matching of Organ Donors and Recipients using Big Data
Project Overview
Organ transplantation represent potentially enormous improvements in quality-of-life -- even life-saving treatment -- for patients of various chronic diseases, but transplantation also entails many risks and complications that can outweigh the benefits. Many of these risks and complications could be greatly reduced or avoided entirely with accurate pre-operative assessment of the match between donor and recipient. Surprisingly, for most organ transplantations, no generally-accepted methods of assessment and matching are currently in use. (Different transplantation programs -- sometimes different surgeons within the same transplantation program -- typically use different rules-of-thumb.) This project develops novel machine learning algorithms that adaptively adjust complexities in order to learn (to a very fine granularity) the complex and subtle interactions between the clinical traits of donors and recipients. These algorithms are applicable to a wide range of organ transplants -- from kidneys to hearts to corneas -- and yield greatly improved matches between donors and recipients, more efficient surgical decisions and significantly improved clinical outcomes.
Principal Investigator
-
Prof. Mihaela van der Schaar
Students
Collaborators