Ó³»­´«Ã½

In collaboration with the Machine Learning for Health team, researchers in CVDi construct, train, and use machine learning models to deeply analyze large scale diagnostic imaging and ECG datasets. When paired with patient clinical records, this work aids identification and prediction of cardiovascular disease diagnosis, which can help identify at-risk individuals who may benefit from preventive care. When combined with genetic association studies, these data uncover insights into the genetic architecture of different forms of cardiovascular disease. 

A few examples of how the CVDi team has made use of machine learning so far include:

• Analysis of ECG datasets to enable scalable cardiovascular screening and risk stratification for atrial fibrillation and heart failure
• Identification of mitral valve prolapse and prediction of clinical endpoints using echocardiogram imaging
• Leveraging machine learning and genetic association studies to quantify fibrosis in MRI data across multiple organs to identify genes and pathways that are common across multiple organs