A Multimodal System for Comprehensive Cardiovascular Monitoring Using ECG, PCG, and PPG Signal Fusion

Figure 1. The simplified functional block diagram of the proposed vital-sign monitoring system comprises the analog circuits for ECG, PCG, and PPG acquisition, as well as the digital components, including the microcontroller and power supply.

Figure 2. Photograph of the proposed system.

Figure 3. (a) Experimental protocol. (b) Experimental setup (left) and sensor placement on the human body (right).

Figure 4. Workflow of digital signal processing for ECG, PCG, and PPG data. Each signal type is segmented and filtered prior to feature extraction. Extracted features include both individual signal-based features and combined features derived from pairs of signals.

Figure 5. (a) Example of segmented and filtered ECG, PCG, and PPG signals stacked together, with the peak detection algorithm applied. (b) Example of how PEP, PAT, and PTT are related.

Figure 6. Distributions of R–R interval (RRI) from ECG, left ventricular ejection time (LVET) from PCG, and crest time from PPG, computed from segmented recordings and aggregated across all participants.

Figure 7. Correlation of RR interval, an ECG-derived heart rate variability metric, with age and BMI.

Figure 8. Correlation of LVET, a PCG-derived index of systolic function, with age and BMI.

Figure 9. Correlation of crest time, a PPG-derived measure of arterial compliance and pulse dynamics, with age and BMI.

Figure 10. Relationships between PEP with age and BMI.

Figure 11. Relationships between PTT with age and BMI.

Figure 12. Relationships between PAT with age and BMI.