Predictive Modeling of Acute Cardiovascular Events Using Continuous Wearable Sensor Data and Deep Learning Architectures
Keywords:
Cardiovascular disease; Wearable technology; Deep learning; Long Short-Term Memory (LSTM); Predictive analytics; Digital health; Remote patient monitoring.Abstract
Cardiovascular diseases remain a leading cause of global mortality, making the early detection of acute cardiovascular events (ACEs) a critical clinical priority. While wearable technology provides unprecedented access to continuous physiological data, effectively translating this data into actionable clinical predictions remains a challenge. This study aims to develop and validate a deep learning framework for predicting ACEs up to 48 hours in advance using multimodal wearable sensor data. We conducted a retrospective cohort study utilizing continuous heart rate, heart rate variability (HRV), and blood oxygen saturation (SpO2) data collected from 15,420 patients over a 12-month period. A Long Short-Term Memory (LSTM) neural network was trained and compared against traditional machine learning baselines, including Logistic Regression and Random Forest. The proposed LSTM model achieved an Area Under the Receiver Operating Characteristic Curve (AUROC) of 0.89, significantly outperforming the baseline models (AUROC of 0.74 and 0.78, respectively). Furthermore, the model demonstrated a sensitivity of 85.2% and a specificity of 82.4% in the 48-hour prediction window. These findings indicate that deep learning applied to continuous, non-invasive wearable data holds substantial potential for the early prediction of acute cardiovascular events, paving the way for proactive, personalized clinical interventions.
Downloads
References
Perez MV, Mahaffey KW, Hedlin H, Rumsfeld JS, Garcia A, et al. Large-scale assessment of a smartwatch to identify atrial fibrillation. N Engl J Med. 2019;381(20):1909-1917.
(Supports the use of wearable technology for large-scale cardiovascular monitoring and event detection).
Hannun AY, Rajpurkar P, Haghpanahi M, Tison GH, Bourn C, et al. Cardiologist-level arrhythmia detection and classification in ambulatory electrocardiograms using a deep neural network. Nat Med. 2019;25(1):65-69.
(Validates the application of deep learning architectures for analyzing continuous, ambulatory physiological data).
Shickel B, Tighe PJ, Bihorac A, Rashidi P. Deep EHR: a survey of recent advances in deep learning techniques for electronic health record (EHR) analysis. IEEE J Biomed Health Inform. 2018;22(5):1589-1604.
(Provides foundational context for using Long Short-Term Memory (LSTM) networks and other deep learning models on sequential/time-series health data).
Jandackova VK, Britton A, Malik M, Steptoe A. Heart rate variability and risk of cardiovascular events in healthy men and women: a prospective cohort study. Int J Epidemiol. 2016;45(4):1169-1177.
(Supports the clinical premise that Heart Rate Variability (HRV) is a critical predictive biomarker for acute cardiovascular events).
Topol EJ. High-performance medicine: the convergence of human and artificial intelligence. Nat Med. 2019;25(1):44-56.
(Offers a broad, authoritative perspective on how AI and continuous remote patient monitoring are transforming proactive cardiovascular care).
