Project PrediCT aims to increase the predictive power of NCCT CAC (coronary artery calcium) scans for major adverse cardiac events (MACE) by integrating other features alongside calcium formation and enhancing the NCCT scan itself. We are currently focusing on robustly segmenting and visualizing calcium patterns, using the Stanford COCA dataset.
Building and Comparing Segmentation Strategies for Coronary Artery Calcium (CAC)
Radiomics Feature Extraction and Calcium Phenotype Discovery
Data Augmentation Using Physics-Informed Plaque Growth Simulation