Smartphone screening for atrial fibrillation (AFib), along with causal artificial intelligence (AI) for estimating cardiovascular risk and the use of AI to detect aortic stenosis were the focus of three separate hot line trials presented Aug. 28 as part of ESC Congress 2022 in Barcelona.

In the eBRAVE-AF study of 5,500 patients in Germany with no known AFib who were randomly assigned to digital screening or usual care, researchers reported that digital screening using ordinary smartphones more than doubled the detection rate of treatment-relevant AFib among a broad elderly target population between the ages of 50 and 90.

"Anyone with a smartphone can screen [themselves] for the world's most important cardiac arrhythmia," said Axel Bauer, MD, who presented the findings. "This might have huge implications for prevention of stroke; however, future studies are needed to test whether improved AFib diagnostics through digital technologies translate into better treatment outcomes."

In a second study presented by Brian A. Ference, MD, MPhil, MSc, FACC, causal AI was found to substantially improve the validity of estimating cardiovascular risk and benefits. Ference noted that current risk estimating algorithms do not include the causal effects and can lead to "a series of counter-intuitive and biologically implausible conclusions." Based on the findings from his study, Ference said that embedding causal effects into risk estimating algorithms addresses these challenges and accurately estimates baseline cardiovascular risk and benefits starting at any age and extending for any duration.

"Causal AI produces AI algorithms that can accurately predict risk and for the first time also prescribe specific actions to reduce risk by accurately estimating expected benefit," he said. "Thus, creating a new generation of AI algorithms that can be used to guide individual treatment decisions to personalize the prevention of cardiovascular disease."

In a third presentation, Geoffrey A. Strange, PhD, discussed results of the AI-ENHANCED study, which he said demonstrated that "an AI decision-support algorithm can automatically identify patients with moderate-to-severe forms of aortic stenosis generally associated with poor survival if left untreated." According to Strange, AI decision-support correctly identified five-year mortality in 56% of patients with moderate aortic stenosis and in 67% of those with severe aortic stenosis.

Clinical Topics: Arrhythmias and Clinical EP, Valvular Heart Disease, Atrial Fibrillation/Supraventricular Arrhythmias

Keywords: ESC Congress, ESC22, ACC International, Innovation, Atrial Fibrillation, Smartphone, Artificial Intelligence, Digital Technology, Aortic Valve Stenosis, Stroke

< Back to Listings