Researchers have developed an artificial intelligence program called PanEcho that shows remarkable accuracy in interpreting echocardiograms, according to findings presented at the American Heart Association's Scientific Sessions 2024. This breakthrough could significantly reduce wait times for heart imaging results and lead to more timely medical interventions.
PanEcho, developed by the Cardiovascular Data Science Lab at Yale School of Medicine, is the first AI system capable of automatically assessing all key areas of heart health from echocardiograms using images from multiple views. The program demonstrated high accuracy across 18 different diagnostic classification tasks, with an average score of 0.91 on the area under the receiver operating characteristic curve (AUC), where 1.0 represents perfect accuracy.
Gregory Holste, M.S.E., a researcher with the Yale School of Medicine, highlighted PanEcho's potential for use in simplified, AI-assisted screening echocardiograms. 'In settings where expert readers may not be readily accessible, PanEcho could rapidly rule out abnormalities that would otherwise require urgent referral,' Holste explained.
The AI program showed particularly impressive results in assessing ventricle function and structure, with AUC scores ranging from 0.91 to 0.98 for various conditions. It also excelled in diagnosing valvular diseases, achieving AUC scores between 0.93 and 0.99 for conditions such as aortic stenosis and mitral regurgitation.
PanEcho's ability to estimate continuous echocardiographic parameters was also noteworthy, with a median normalized mean absolute error of 0.13 across 21 tasks. This accuracy extended to critical measurements of left ventricle dimensions and function, which are essential for assessing overall heart health.
While these results are promising, the researchers acknowledge that further validation in real-world clinical settings is necessary. The next steps include evaluating PanEcho's performance with portable echocardiogram machines used in emergency rooms and smaller medical clinics, where AI tools could have the most significant impact.
This advancement in AI for echocardiography represents a significant step forward in cardiac diagnostics. If successfully implemented in clinical practice, PanEcho could potentially improve access to timely cardiac care, especially in areas with limited access to specialist readers. The technology's ability to quickly and accurately interpret complex heart images could lead to faster diagnoses and more efficient treatment planning, ultimately benefiting patients with various heart conditions.
