Related Experiment Video
Updated: Sep 29, 2025
05:03
Patient Directed Recording of a Bipolar Three-Lead Electrocardiogram using a Smartwatch with ECG Function
Published on: December 11, 2019
8.8K
Deep Learning Forecasts the Occurrence of Sleep Apnea from Single-Lead ECG
Mahsa Bahrami1, Mohamad Forouzanfar2,3,4
1Department of Biomedical Engineering, K. N. Toosi University of Technology, Tehran, Iran.
Cardiovascular Engineering and Technology
|March 18, 2022
Summary
Accurate sleep apnea forecasting is now possible using deep learning on single-lead ECG data. This breakthrough enables early detection and management of sleep apnea, improving patient outcomes.
Area of Science:
- Biomedical Engineering
- Artificial Intelligence in Medicine
- Cardiology
Background:
- Sleep apnea is a prevalent sleep disorder with significant health risks if untreated.
- Early forecasting of apnea events is crucial for timely intervention and management.
Purpose of the Study:
- To develop and validate a novel framework for forecasting sleep apnea occurrence.
- To leverage deep recurrent neural networks (DRNNs) for apnea prediction using electrocardiogram (ECG) data.
Main Methods:
- Extracted and aligned ECG R-peak amplitudes and R-R intervals using power spectral analysis.
- Developed recurrent deep learning models to identify predictive ECG features.
- Forecasted apnea occurrence up to five minutes in advance.
Main Results:
- Achieved a forecasting accuracy of up to 94.95% for apnea events.
- Demonstrated superior performance compared to conventional multilayer perceptron and other state-of-the-art techniques (p < 0.05).
Conclusions:
- The proposed deep learning approach effectively forecasts sleep apnea from single-lead ECG.
- This method is suitable for integration into wearable sleep monitors for sleep apnea management.
- The developed algorithms are publicly available on GitHub.
