“Developing Artificial Intelligence Models to Read Electrocardiograms and Predict the Risk of Stroke and Heart failure”

Project description

Undetected atrial fibrillation (AF) is a major health concern with life-threatening complications. Earlier AF diagnosis can only be achieved by more sensitive diagnostic modalities, especially by those employed as screening tools. Automated screening from ECGs coupled with clinical and lab data analysis is a complex task often convoluted by low-quality digitized images. Optimization of such screening tools will enable more intense ECG monitoring and benefit patients. The potential for AI applications in healthcare seems endless, from deep learning algorithms that can diagnose cancer, read CT scans and X-rays better than humans can do to natural language processing applications that can comb through unstructured data in electronic health records and unearth novel and useful information.  This project will adopt a whole system approach to develop Artificial Intelligence (AI) models coupled with image recognition and signal processing to collate information from 12-lead ECGs with clinical and lab data and detect AF in early stages.