The demand for liver transplants exceeds the number of available livers, which leads to approximately 14,000 waitlisted transplant candidates each year. To address this shortage, one approach is to use livers with an acceptable level of high fat content, so more livers are available for transplant. Another approach is to extend the transplant time window of harvested livers by undergoing a freezing and rehabilitation process. This project leverages computer vision techniques, such as image segmentation, and a dataset of biopsy images and pathologist evaluations to aid and assess these approaches. We aim to develop machine learning models to accurately, reliably, and transparently determine livers that have acceptably high fat content and are viable after freezing and rehabilitation.

I am excited to apply the knowledge I have gained from machine learning courses to this research project. This project piques my interest because it uses computer vision to address a problem in healthcare. I hope to learn more about the state of the art in computer vision and methods to improve machine learning model performance. I am participating in SuperUROP because I want to develop high-level research skills and work on a long-term project.