Meeting News

Artificial intelligence, wearable ring device may detect AF

SAN FRANCISCO — A novel wearable smart ring device with a deep learning algorithm detected atrial fibrillation with photoplethysmography monitoring signals, according to data presented at Heart Rhythm Society Annual Scientific Sessions.

“The diagnostic performance is comparable to medical-grade conventional pulse oximeters,” Eue-Keun Choi, MD, PhD, clinical assistant professor in the division of cardiology at Seoul National University Hospital in South Korea, told Cardiology Today. “This technology can be a new strategy to detect AF in the future.”

Soonil Kwon, MD, of the department of internal medicine at Seoul National University Hospital in South Korea, and colleagues analyzed data from 119 patients with AF who underwent simultaneous ECG and photoplethysmography before and after direct-current cardioversion.

Photoplethysmography data were analyzed through a deep learning algorithm that was developed with a convolutional neural network. Rhythms were then interpreted with a novel noninvasive ring-type wearable device that used photoplethysmography signals based on the convolutional neural network.

“Deep learning or [artificial intelligence] can overcome formerly important problems of [photoplethysmography]-based arrhythmia diagnosis,” Choi said in an interview. “It not only improves diagnostic accuracy in great degrees, but also suggests a metric how this diagnosis will be likely true without ECG validation. Combined with wearable technology, this will considerably boost the effectiveness of AF detection.”

Throughout the study, there were 27,569 photoplethysmography samples. The accuracy of the convolutional neural network was 99.3% to diagnose AF and 95.9% to diagnose sinus rhythm. The performance of this neural network was superior to previously developed algorithms, according to the abstract.

The neural network was able to accurately diagnose sinus rhythm even with a high burden of premature atrial complexes.

The average confidence level was 98.6% for true classification and 80.5% for false classification.

A novel wearable smart ring device with a deep learning algorithm detected atrial fibrillation with photoplethysmography monitoring signals, according to data presented at Heart Rhythm Society Annual Scientific Sessions.
Source: Adobe Stock

The diagnostic accuracy of the wearable device was 98.3% for sinus rhythm and 100% for AF after filtering low-quality samples.

“We would like to evaluate the deep learning algorithm with a newly developed ring device in daily activity,” Choi told Cardiology Today. “This will provide feasibility for AF screening in a high-risk population. Also, we hope that the ring device could be used for AF detection in a clinical trial due to its noninvasiveness.” – by Darlene Dobkowski

Reference:

Kwon S, et al. Abstract S-SP06-05. Presented at: Heart Rhythm Society Annual Scientific Sessions; May 8-11, 2019; San Francisco.

Disclosures: Kwon reports no relevant financial disclosures. Choi reports he received a research grant from Biosense Webster, Bristol-Myers Squibb/Pfizer, Daiichi-Sankyo and Sky Labs and is a stockholder of Sky Labs.

SAN FRANCISCO — A novel wearable smart ring device with a deep learning algorithm detected atrial fibrillation with photoplethysmography monitoring signals, according to data presented at Heart Rhythm Society Annual Scientific Sessions.

“The diagnostic performance is comparable to medical-grade conventional pulse oximeters,” Eue-Keun Choi, MD, PhD, clinical assistant professor in the division of cardiology at Seoul National University Hospital in South Korea, told Cardiology Today. “This technology can be a new strategy to detect AF in the future.”

Soonil Kwon, MD, of the department of internal medicine at Seoul National University Hospital in South Korea, and colleagues analyzed data from 119 patients with AF who underwent simultaneous ECG and photoplethysmography before and after direct-current cardioversion.

Photoplethysmography data were analyzed through a deep learning algorithm that was developed with a convolutional neural network. Rhythms were then interpreted with a novel noninvasive ring-type wearable device that used photoplethysmography signals based on the convolutional neural network.

“Deep learning or [artificial intelligence] can overcome formerly important problems of [photoplethysmography]-based arrhythmia diagnosis,” Choi said in an interview. “It not only improves diagnostic accuracy in great degrees, but also suggests a metric how this diagnosis will be likely true without ECG validation. Combined with wearable technology, this will considerably boost the effectiveness of AF detection.”

Throughout the study, there were 27,569 photoplethysmography samples. The accuracy of the convolutional neural network was 99.3% to diagnose AF and 95.9% to diagnose sinus rhythm. The performance of this neural network was superior to previously developed algorithms, according to the abstract.

The neural network was able to accurately diagnose sinus rhythm even with a high burden of premature atrial complexes.

The average confidence level was 98.6% for true classification and 80.5% for false classification.

A novel wearable smart ring device with a deep learning algorithm detected atrial fibrillation with photoplethysmography monitoring signals, according to data presented at Heart Rhythm Society Annual Scientific Sessions.
Source: Adobe Stock

The diagnostic accuracy of the wearable device was 98.3% for sinus rhythm and 100% for AF after filtering low-quality samples.

“We would like to evaluate the deep learning algorithm with a newly developed ring device in daily activity,” Choi told Cardiology Today. “This will provide feasibility for AF screening in a high-risk population. Also, we hope that the ring device could be used for AF detection in a clinical trial due to its noninvasiveness.” – by Darlene Dobkowski

Reference:

Kwon S, et al. Abstract S-SP06-05. Presented at: Heart Rhythm Society Annual Scientific Sessions; May 8-11, 2019; San Francisco.

Disclosures: Kwon reports no relevant financial disclosures. Choi reports he received a research grant from Biosense Webster, Bristol-Myers Squibb/Pfizer, Daiichi-Sankyo and Sky Labs and is a stockholder of Sky Labs.

    See more from Heart Rhythm Society Scientific Sessions