0
0
Atrial fibrillation (AFib), the most common cardiac arrhythmia globally affecting approximately 59 million people in 2019, poses significant risks of heart failure, dementia, and stroke. Detecting and treating AFib early is paramount, and researchers from the Luxembourg Centre for Systems Biomedicine (LCSB) at the University of Luxembourg have made a groundbreaking advancement in this area.
Published in the scientific journal Patterns, the study unveils a deep-learning model capable of predicting the transition from a normal cardiac rhythm to AFib. This predictive model offers early warnings, on average 30 minutes before AFib onset, with an impressive accuracy of around 80%. These findings open avenues for integration into wearable technologies, promising better patient outcomes through early interventions.
“AFib imposes a considerable burden on healthcare systems, underscoring the urgency for early detection and treatment,” says Prof. Jorge Goncalves, head of the Systems Control group at LCSB. “Our model, named WARN (Warning of Atrial fibRillatioN), marks a significant departure from conventional approaches by prospectively predicting AFib onset.”
The WARN model utilizes heart rate data to identify different cardiac phases and calculate the probability of an imminent AFib episode. Trained and tested on 24-hour recordings from 350 patients at Tongji Hospital (Wuhan, China), WARN demonstrated remarkable efficacy in providing early warnings, enabling timely interventions.
“An exciting aspect of our model is its reliance on readily available heart rate data, making it compatible with wearable devices like smartwatches,” explains Dr. Marino Gavidia, lead author of the study. “This paves the way for real-time monitoring and early warnings from convenient wearable devices.”
Moreover, the model’s low computational cost makes it suitable for integration into smartphones, facilitating seamless data processing from smartwatches. Patients could benefit from continuous cardiac rhythm monitoring, receiving early warnings to initiate preventive measures like antiarrhythmic medication.
“Our ultimate goal is to empower patients with personalized models, allowing for continuous refinement based on daily heart dynamics,” adds Prof. Goncalves. “This iterative approach holds promise for reducing emergency interventions and improving patient outcomes.”
As the research progresses, the team aims to conduct clinical trials and explore innovative therapeutic interventions, leveraging the potential of deep-learning models to revolutionize cardiac care.
The groundbreaking work by LCSB researchers marks a significant stride toward proactive management of AFib, offering hope for millions worldwide battling this prevalent cardiac condition.
