Cedars-Sinai AI Tool Accurately Assesses, Diagnoses Cardiac Function
A Cedars-Sinai artificial intelligence tool outperformed ultrasound technicians at assessing and diagnosing cardiac function in a randomized clinical trial.
Source: Getty Images
By - Researchers from Cedars-Sinai have shown that an artificial intelligence (AI) tool’s performance was superior to that of sonographers when assessing and diagnosing cardiac function, according to the results of a blinded, randomized clinical trial published in Nature this week.
The work builds on previous research undertaken by teams at Cedars-Sinai and Stanford University to develop an AI tool to assess one of the key metrics used to diagnose cardiac function: left ventricular ejection fraction.
In this non-inferiority trial, the researchers intended to demonstrate that the tool did not perform worse than sonographers, also known as ultrasound technicians, at assessing and diagnosing cardiac function using electrocardiogram images.
To evaluate performance, the researchers tasked the tool and 25 sonographers with assessing images from a group of 3,495 transthoracic echocardiogram studies performed between June 1, 2019 and August 8, 2019.
The included image studies were then split between the AI and the sonographers, with 1,740 studies assigned to the AI group and 1,755 studies assigned to the sonographer group. Following these initial evaluations, 10 cardiologists performed a blinded assessment of each evaluation.
The researchers found that the cardiologists were unable to tell which assessments were made by sonographers and which were made by AI, but more frequently agreed with the AI’s assessment of the imaging.
Further, the cardiologists made corrections to 16.8 percent of the initial assessments made by AI, compared to 27.2 percent of those made by the sonographers.
“[These findings speak] to the strong performance of the AI algorithm as well as the seamless integration into clinical software,” said cardiologist David Ouyang, MD, principal investigator of the clinical trial and senior author of the study, in a press release published by Cedars-Sinai. “We believe these are all good signs for future AI trial research in the field.”
The researchers also indicated that they believe the tool may save clinicians time and streamline the cardiac imaging workflow when it is implemented at Cedars-Sinai and other US health systems. The press release does not denote when such an implementation will take place or what Cedar-Sinai’s plans for implementation are.
The clinical trial also helps set a precedent for how clinical AI algorithms are developed and tested within health systems, which has the potential to support deployment efforts and improve patient care, explained Sumeet Chugh, MD, director of the Division of Artificial Intelligence in Medicine and the Pauline and Harold Price Chair in Cardiac Electrophysiology Research at Cedars-Sinai, in the press release.
“This work raises the bar for artificial intelligence technologies being considered for regulatory approval, as the Food and Drug Administration has previously approved artificial intelligence tools without data from prospective clinical trials,” said Susan Cheng, MD, MPH, director of the Institute for Research on Healthy Aging in the Department of Cardiology at the Smidt Heart Institute and co-senior author of the study. “We believe this level of evidence offers clinicians extra assurance as health systems work to adopt artificial intelligence more broadly as part of efforts to increase efficiency and quality overall.”
