Automated Fatty Liver Diagnosis

Up to 30% of the population has a Fatty Liver Disease (FLD, a condition in which fat builds up in your liver). Non-invasive ultrasound assessment of this liver condition is an increasing demand in healthcare service due to its high risks leads to advanced liver diseases. However, an ultrasound-based examination has made the manual inspection a lengthy and tedious task and observer dependent. The proposed research aims at a computer-aided liver ultrasound assessment software toolkit facilitating the diagnosis of FLD. In particular, this proposal will design computational methods, such as machine learning and recent deep learning, to automatically extract related features in ultrasound data, and to detect and grade the levels of FLD. This project is a pioneering attempt. In both industry and academia, there is limited prior work. It will provide a strategic base towards a full automatic liver ultrasound diagnosis system for the partner organization.

Faculty Supervisor:

Shuo Li

Student:

Zhifan Gao

Partner:

Medo

Discipline:

Medicine

Sector:

University:

Program:

Accelerate