Utilizing AI and Attention to Assist in Training Medical Students to Read and Diagnose Electrocardiograms

Electrocardiograms allow for the diagnosis and analysis of cardiovascular conditions. Despite this, there is a lack of consistent analytical strategies. This results in newer medical practitioners, often students, fixating on irrelevant segments. This leads to frequent misdiagnosis and a high level of inconsistencies. By using machine learning strategies and analyzing human attention networks, recorded through the medium of eye-tracking, we hope to develop a tool to aid medical students and medical practitioners in making consistent and accurate electrocardiogram diagnoses.