Multiple Physiological Signals Fusion Techniques for Improving Heartbeat Detection: A Review

Area of Science:

• Biomedical Engineering
• Signal Processing

Background:

• Robust heartbeat detection is crucial for various medical applications.
• Multi-modal physiological signals offer complementary information for improved accuracy.

Purpose of the Study:

• To review and synthesize techniques for robust heartbeat detection using fused physiological signals.
• To identify optimal signal combinations and fusion strategies.

Main Methods:

• Literature review of multi-modal signal fusion techniques for heartbeat detection.
• Analysis of preprocessing steps including filtering, normalization, and resampling.
• Examination of fusion approaches: RR interval exploitation, signal selection, voting, and advanced methods (Bayesian, HMM, neural networks).

Main Results:

• Electrocardiogram (ECG), blood pressure (BP), and artificial blood pressure (ABP) are primary signals for robust detection.
• Inclusion of stroke volume (SV), photoplethysmogram (PPG), electromyogram (EMG), electroencephalogram (EEG), and electrooculogram (EOG) can further enhance performance.
• Signal quality assessment and delay correction are critical preprocessing steps.

Conclusions:

• Multi-modal signal fusion significantly improves heartbeat detection robustness.
• Guidelines for future research and performance comparison are provided.
• Future research should focus on advanced fusion techniques and signal quality assessment.