Research highlights:
-
A new software program that uses artificial intelligence to read echocardiograms could shorten wait times for results and lead to timely medical care.
-
PanEcho is the first AI system to automatically assess all key areas of heart health from multi-view echocardiograms and identify which views are most relevant to each imaging task.
-
Please note: the research in this press release is a research summary. Abstracts presented at the American Heart Association scientific meetings are not peer-reviewed and the findings are considered preliminary until published as full manuscripts in a peer-reviewed scientific journal.
Embargoed until 3:55 PM CT/4:55 PM ET, Saturday, November 16, 2024
(NewMediaWire) – November 16, 2024 – CHICAGO Using an artificial intelligence program to read echocardiograms could shorten the wait time for results and help lead to timely medical care, according to the latest science presented today at the American Heart meeting Association Scientific sessions 2024. The meeting, November 16-18, 2024, in Chicago, is a premier global exchange of the latest scientific advances, research and evidence-based clinical practice updates in cardiovascular science.
In this study, an artificial intelligence program called PanEcho was tested for its ability to independently interpret echocardiography videos. An echocardiogram is a type of heart imaging that allows doctors to see the structure and function of the heart, and it is used to diagnose and treat heart disease.
PanEcho builds on previous AI applications in cardiology that were limited to single views of the heart and disease-specific criteria. The research team developed a new AI system capable of comprehensive reporting of all key findings from each set of echocardiography videos.
“PanEcho has the potential to be used in simplified, AI-assisted screening echocardiograms,” said Gregory Holste, MSE, researcher at the Cardiovascular Data Science (CardS) Lab at the Yale School of Medicine in New Haven, Connecticut, who developed the presentation presented. study. “In settings where expert readers may not be readily accessible, PanEcho could quickly rule out abnormalities that would otherwise require urgent referral.”
PanEcho’s diagnostic performance was evaluated using a standard diagnostic test accuracy measure: the area under the receiver operating characteristic curve (AUC). A 100% accurate test has an AUC of 1, and an uninformative test (for example, random guessing) has an AUC of 0.5.
The study found:
-
When PanEcho’s capabilities were evaluated in 18 different diagnostic classification tasks, the average score was 0.91.
-
When assessing ventricular function and structure, PanEcho accuracy scores were:
-
0.95 AUC to detect increased size in the left ventricle, making it difficult for the heart to pump blood.
-
0.98 AUC to identify systolic dysfunction in the left ventricle reduces the ability of the chamber to pump blood into the aorta.
-
0.91 AUC to detect left ventricular hypertrophy, a term used when the heart’s left pumping chamber has become thickened and may not be pumping efficiently.
-
0.93 AUC to identify systolic dysfunction in the right ventricle reduces the ability of the chamber to pump blood into the lungs.
-
In diagnosing valve disease, PanEcho accuracy scores were:
-
0.99 AUC to identify severe aortic stenosis, restricted blood flow due to calcium buildup on the heart valves.
-
0.96 AUC to identify mitral stenosis, narrowing of the mitral valve between the left atrium and the left ventricle.
-
0.93 AUC to identify moderate or severe aortic regurgitation, leakage of blood through the aortic valve.
-
0.96 AUC to identify moderate or greater mitral regurgitation, leakage of blood posteriorly through the mitral valve.
PanEcho was also evaluated for its ability to estimate continuous echocardiographic parameters, using mean absolute error, which is a measurement of the average distance between predicted values and actual values, meaning that the smaller the distance, the more accurate the prediction. When evaluated in 21 tasks, PanEcho had a median normalized mean absolute error of 0.13.
PanEcho showed accuracy in quantifying left ventricular size and function, including:
-
4.4% mean absolute error in estimating left ventricular ejection fraction;
-
1.3 mm mean absolute error in estimating the thickness of the intraventricular septum of the left ventricle; And
-
A mean absolute error of 1.2 mm in estimating the posterior wall thickness of the left ventricle.
These measurements are critical for accurately assessing the structure and function of the left ventricle, an important aspect of heart health.
“This work represents an advancement in the field of AI for echocardiography, and we hope that the public release of our AI model will encourage the research community to move to flexible, multi-task, multi-view approaches for echocardiogram interpretation,” added he added.
PanEcho is limited by its retrospective validation in previously acquired data. The next step is to prospectively validate its application in real patient care settings to provide further insight into its clinical viability, Holste said. “It should also be evaluated for use with portable echocardiogram machines used in emergency rooms and smaller medical clinics, where there is the potential for the greatest positive impact with AI tools.”
Study background, design and details:
-
PanEcho was developed using 1.23 million multi-view echocardiogram videos that were part of nearly 34,000 transthoracic echocardiography tests performed for people treated at a Yale-New Haven Health System hospital (five locations in Connecticut) or who were seen in one of the system’s outpatient clinics. between 2016 and 2022. The AI program was developed by the CarDS Lab at the Yale School of Medicine.
-
The echocardiograms came from 26,067 unique individuals whose routine care included this type of imaging.
-
Approximately 52% of the image data collected came from adults who self-identified as male.
-
The average age of the people whose echocardiograms were part of this collection was 67 years.
-
Approximately 80% of the participants in the image collection were noted as white people, 14.2% were black people, 1.8% were Asian people, and 4.3% were people who self-identified as another race in this study.
The principal investigator of this study was Rohan Khera, MD, MS, head of the CarDS Lab at Yale School of Medicine. The work was led by co-first authors and CarDS Lab members, Gregory Holste, MSE, and Evangelos Oikonomou, MD, D.Phil. Additional co-authors, disclosures, and funding sources are noted in the abstract.
Statements and conclusions of studies presented at American Heart Association scientific meetings are solely those of the study authors and do not necessarily reflect the policy or position of the association. The Association makes no representation or warranty as to its accuracy or reliability. Abstracts presented at the association’s scientific meetings are not peer-reviewed, but are compiled by independent review panels and are judged on the basis of their potential to contribute to the diversity of scientific issues and views discussed at the meeting . The findings are considered preliminary until published as a full manuscript in a peer-reviewed scientific journal.
The association receives funding mainly from private individuals; foundations and corporations (including pharmaceutical companies, device manufacturers, and other companies) also make donations and fund specific association programs and events. The association has a strict policy to prevent these relationships from influencing scientific content. Earnings from pharmaceutical and biotech companies, device manufacturers and health insurers and the association’s general financial information are available here.
Additional Resources:
###
About the American Heart Association
The American Heart Association is a relentless force for a world of longer, healthier lives. We are committed to ensuring health equity in all communities. Working with countless organizations, and powered by millions of volunteers, we fund innovative research, advocate for public health, and share lifesaving resources. The Dallas-based organization has been a leading source of health information for a century. In 2024, our centenary, we celebrate our rich 100-year history and achievements. As we enter our second century of bold discovery and impact, our vision is to advance health and hope for everyone, everywhere. Contact us at heart.org, Facebook, X or by calling 1-800-AHA-USA1.
For media inquiries and AHA expert perspectives:
AHA Communications and Media Relations in Dallas: 214-706-1173; ahacommunications@heart.org
Bridgette McNeill: bridgette.mcneill@heart.org
For public inquiries: 1-800-AHA-USA1 (242-8721)
heart.org and stroke.org
