Researchers have developed a new artificial intelligence program that could revolutionize echocardiogram analysis, potentially leading to faster diagnoses and more timely medical interventions. The program, named PanEcho, was presented at the American Heart Association's Scientific Sessions 2024 in Chicago as part of late-breaking science. PanEcho is designed to interpret echocardiography videos independently, assessing multiple aspects of heart health from various imaging views. This capability sets it apart from previous AI applications in cardiology, which were often limited to single views of the heart and specific disease criteria.
Gregory Holste, M.S.E., a researcher with the Cardiovascular Data Science (CarDS) Lab at the Yale School of Medicine, presented the study findings. Holste emphasized the potential of PanEcho to be used in simplified, AI-assisted screening echocardiograms, particularly in settings where expert readers may not be readily available. The program's performance was evaluated using the area under the receiver operating characteristic curve (AUC), a standard measure of accuracy for diagnostic tests. Across 18 different diagnostic classification tasks, PanEcho achieved an impressive average score of 0.91, with 1.0 representing perfect accuracy.
PanEcho demonstrated high accuracy in assessing various aspects of heart function and structure. For instance, it achieved a 0.95 AUC for detecting increased size in the left ventricle, a 0.98 AUC for identifying systolic dysfunction in the left ventricle, and a 0.99 AUC for identifying severe aortic stenosis. The AI program also showed promise in estimating continuous echocardiographic parameters. It achieved a 4.4% mean absolute error when estimating left ventricle ejection fraction, a critical measure of heart function.
PanEcho was developed using a vast dataset of 1.23 million echocardiogram videos from nearly 34,000 transthoracic echocardiography tests conducted at Yale-New Haven Health System hospitals and outpatient clinics between 2016 and 2022. The diverse dataset included imaging from over 26,000 unique individuals with an average age of 67. While the results are promising, Holste noted that the next step is to validate PanEcho's application in real-world patient care environments. This prospective validation will provide further insights into its clinical viability.
Additionally, researchers aim to evaluate its use with portable echocardiogram machines used in emergency rooms and smaller medical clinics, where AI tools could have the most significant positive impact. The development of PanEcho represents a significant advance in AI for echocardiography. The research team hopes that the public release of their AI model will encourage the broader research community to move towards flexible, multi-task, multi-view approaches for echocardiogram interpretation. As AI continues to make inroads in medical diagnostics, tools like PanEcho have the potential to enhance the speed and accuracy of heart health assessments, ultimately leading to improved patient outcomes and more efficient healthcare delivery.
