Summer Research Project Grant Report
Cross-disciplinary Domains: Artificial Intelligence, Cardiology, Medical Imaging
Research Project Objective
- Develop a novel zero-shot medical segmentation system.
- Build a pipeline using the system for the automatic diagnosis of steatotic liver disease (SLD) from cardiac computed tomography (CCT) scans.
Background
Cardiovascular diseases (CVDs) remain the leading cause of death worldwide [1]. Recent studies suggest a strong link between liver and heart health [2], motivating automated methods for assessing liver health. The liver is the body’s largest internal organ, performing essential functions including metabolic regulation, detoxification, bile production, and nutrient storage. Its central role in lipid metabolism makes it particularly susceptible to fat accumulation disorders. Steatotic liver disease (SLD) is characterised by excessive fat accumulation in liver cells, encompassing a spectrum from simple steatosis to more severe forms such as fibrosis and cirrhosis. SLD is highly prevalent, often asymptomatic, and is associated with increased cardiovascular risk. Assessing SLD from CCT scans could therefore provide a more comprehensive picture of cardiometabolic risk.
