A novel ECG-TMS approach for accurate heart rate measurement during sleep using wavelet-based signal separation

Accurate heart rate (HR) measurement during sleep is critical for diagnosing cardiovascular and sleep-related disorders. However, traditional methods relying on standalone heart rate monitors or accelerometers often struggle to differentiate HR changes caused by physiological conditions from those induced by physical movements. This study introduces the ECG-Trunk Muscle Signal (ECG-TMS) method, a novel approach that enhances HR measurement accuracy by isolating muscle artifacts from cardiac signals using discrete wavelet transform (DWT). Data were collected from a healthy male subject under simulated sleep conditions using a Shimmer3 ECG unit. The proposed algorithm successfully separates trunk muscle activity from raw ECG signals, significantly improving the signal-to-noise ratio (SNR) and enabling precise HR analysis even during subtle movements. Compared to accelerometer-based methods, ECG-TMS demonstrates superior sensitivity to minor muscle contractions and robust artifact removal, ensuring reliable HR tracking across varying conditions. These findings highlight the method’s potential for applications in sleep studies, cardiovascular health monitoring, and wearable technology. Future work will focus on real-world validation and algorithm optimization for real-time use, paving the way for a transformative tool in clinical and research settings.