Automatic Detection of QRS Complexes Using Dual Channels Based on U-Net and Bidirectional Long Short-Term Memory

Area of Science:

• Biomedical Engineering
• Signal Processing
• Artificial Intelligence in Medicine

Background:

• Electrocardiogram (ECG) signal analysis is crucial for cardiac health evaluation.
• Accurate QRS complex detection is vital for preliminary cardiac diagnosis.
• Noise and complex signal morphologies challenge automatic QRS detection reliability.

Purpose of the Study:

• To develop an accurate and reliable algorithm for automatic QRS complex detection in ECG signals.
• To address the limitations of existing methods in handling noise and signal variability.

Main Methods:

• A novel dual-channel algorithm combining U-Net and bidirectional long short-term memory (BiLSTM) was proposed.
• A preprocessing step involving mean filtering and discrete wavelet transform was employed for noise reduction.
• Signal transformation and annotation relabeling were performed prior to detection.

Main Results:

• The algorithm achieved high performance on the MIT-BIH arrhythmia database and CPSC2019 dataset.
• Sensitivity reached 99.06% and 95.13%, positive predictivity was 99.22% and 82.03%, and accuracy was 98.29% and 78.73% respectively.
• The results demonstrate the effectiveness of the proposed method for QRS complex detection.

Conclusions:

• The proposed U-Net and BiLSTM-based dual-channel method offers a robust solution for automatic QRS complex detection.
• The algorithm shows significant potential for analyzing large ECG datasets and can be adapted for other medical signal processing tasks.