In comparison to initial sonographer assessment of echocardiograms, cardiologists are over 10 percent less likely to change initial artificial intelligence (AI) assessment of left ventricular ejection fraction (LVEF), according to new research recently presented at the European Society of Cardiology Congress in Barcelona, Spain.
Could artificial intelligence (AI) enhance the quality and efficiency of assessing left ventricular ejection fraction (LVEF) on echocardiograms?
In a recent study involving 3,495 transthoracic echocardiograms, researchers compared sonographer assessment of LVEF versus assessment from a deep learning algorithm (EchoNet-Dynamic), which was reportedly trained on echocardiogram videos and has a documented mean error rate ranging between 4.1 to 6 percent for LVEF assessment.1,2 The study authors sought to determine the frequency of changes greater than 5 percent between the initial LVEF assessment and that of a reviewing cardiologist.
In what the researchers called the first blinded, randomized trial of AI in cardiology, they found that 16.8 percent of the AI assessments were substantially changed by cardiologists in comparison to 27.2 percent of assessments by sonographers.1
“Embedding AI into clinical workflows could potentially provide more precise and consistent evaluations, thereby enabling earlier detection of clinical deterioration or response to treatment,” suggested study co-author David Ouyang, M.D., who is affiliated with the Smidt Heart Institute at Cedars-Sinai in Los Angeles.
Dr. Ouyang, who presented the study findings at the European Society of Cardiology Congress, also emphasized that the reviewing cardiologists largely couldn’t tell if the initial assessment had come from a sonographer or AI assessment. In 1,130 of the cases in the study (43.4 percent), cardiologists said they were unsure if the initial assessment had come from a sonographer or AI. Dr. Ouyang added that cardiologists incorrectly guessed the origin of assessment in 845 cases (24.2 percent).
“We asked our cardiologists over-readers to guess if they thought the tracing they had just reviewed was performed by AI or by a sonographer, and it turns out that they couldn’t tell the difference,” noted Dr. Ouyang. “(This) speaks to the strong performance of the AI algorithm as well as the seamless integration into clinical software.”
References
1. Ouyang D. EchoNet-RCT: blinded, randomized controlled trial of sonographer vs. artificial intelligence assessment of cardiac function. Presented at the European Society of Cardiology Congress: August 27, 2022; Barcelona, Spain.
2. Ouyang D, He B, Ghorbani A, et al. Video-based AI for beat-to-beat assessment of cardiac function. Nature. 2020;580(7802):252-256.
FDA Clears Virtually Helium-Free 1.5T MRI System from Siemens Healthineers
June 26th 2025Offering a cost- and resource-saving DryCool magnet technology, the Magnetom Flow.Ace MRI system reportedly requires 0.7 liters of liquid helium for cooling over the lifetime of the device in contrast to over 1,000 liters commonly utilized with conventional MRI platforms.
Can CT-Based Deep Learning Bolster Prognostic Assessments of Ground-Glass Nodules?
June 19th 2025Emerging research shows that a multiple time-series deep learning model assessment of CT images provides 20 percent higher sensitivity than a delta radiomic model and 56 percent higher sensitivity than a clinical model for prognostic evaluation of ground-glass nodules.