An AI-ECG Algorithm for Left Ventricular Diastolic Dysfunction

Guest: Jae Oh, M.D. 

Host: Anthony H. Kashou, M.D. 

Diastolic function assessment is crucial in diagnosing, managing, and predicting outcomes in various cardiac conditions. It provides insight into heart health, particularly in diagnosing heart failure. Shortness of breath, a common patient complaint, often indicates elevated diastolic filling pressure if linked to a cardiac condition. Echocardiography is the primary method for assessing diastolic function, but it is operator-dependent and not always available. In contrast, ECGs are standardized and widely accessible. Although subtle changes in ECGs are not easily detectable by the human eye, artificial intelligence can identify specific conditions reflected in the ECG. By training an AI model with labeled ECGs based on diastolic function determined through echocardiography, researchers achieved high accuracy in detecting diastolic dysfunction. AI-enhanced ECGs can significantly impact the identification of both asymptomatic and symptomatic cardiac conditions, potentially streamlining diagnostic strategies and reducing costs. Future developments may enable patients to monitor their heart health using simple wearable devices, enhancing the management of heart failure and other conditions.

Topics Discussed:

• Your special clinical academic interest is echocardiography. Why are you interested in ECG AI in diastolic function?

LinkedIn: Mayo Clinic Cardiovascular Services

Podcast episode transcript found here.