Physiological Monitoring Applications of Wearable Multimodal Fusion Systems Based on ECG and PPG: A Comprehensive Review

Figures

• 

  Figure 5. Summary of recent ECG wearable devices: (a) Holter monitors [50]; (b) Mobile cardiac telemetry (MCT) systems [53,55]; (c) ECG monitoring garments [30,46,47,58]; (d) Wrist-based ECG monitors [60,61,62]; (e) Chest-strap ECG monitors [64,65,66,67,68]; (f) Handheld ECG monitors [69,70]; and (g) ECG patch monitors [31,71,72,73,74,75].

• 

  Figure 6. Summary of recent PPG wearable devices: (a) Wrist-worn PPG devices [80,87,88,90]; (b) Finger-based PPG devices (pulse oximeters) [81,91,94,95]; (c) Ear-worn PPG devices [43,82,96,97]; (d) Smartphone camera-based handheld PPG monitoring [98]; (e) Chest-Based PPG devices [102,103,105,106,107,113]; (f) Ring-Based PPG devices [83,84,108,109,110,111]; (g) Head-mounted or forehead PPG devices [112,114,115]; and (h) Arm-worn PPG devices [116,117].

• 

  Figure 1. Schematic representation of a typical electrocardiogram (ECG) waveform illustrating the main components (P wave, QRS complex, and T wave) and clinically relevant time intervals and segments, including the PR interval and segment, ST segment and interval, TP interval, QT interval, and RR interval.

• 

  Figure 2. Schematic representation of a typical photoplethysmography (PPG) waveform illustrating key morphological features, including the systolic peak, dicrotic notch, and diastolic peak. The figure also highlights physiologically relevant temporal parameters such as the peak-to-peak interval, pulse width, and characteristic time differences (ΔT), along with waveform phases associated with atrial depolarization and ventricular repolarization.

• 

  Figure 3. Schematic illustration of the categorization of wearable ECG and PPG monitoring devices according to device type, sensing modality, and anatomical placement.

• 

  Figure 4. Overview of wearable ECG and PPG devices and their integration into multimodal systems. (a) ECG wearable devices; (b) PPG wearable devices; (c) ECG-PPG multimodal wearable systems, including chest-based and wrist-worn configurations.

• 

  Figure 7. Distribution of fusion strategies in ECG-PPG multimodal systems, showing the dominance of feature-level fusion (71.43%) compared to signal-level (21.43%) and decision-level fusion (7.14%).

• 

  Figure 8. Overview of ECG-PPG multimodal fusion pipelines for physiological monitoring applications: (a) Cuffless blood pressure estimation, illustrating ECG and PPG signal acquisition, preprocessing, characteristic point detection, pulse transit time (PTT) calculation, feature extraction, model development and calibration, and systolic and diastolic blood pressure estimation; and (b) Stress monitoring, showing ECG and PPG signal acquisition, signal preprocessing, feature extraction from both modalities, feature fusion, stress classification, and stress level estimation, including binary, ternary, and continuous stress indices.

• 

  Figure 9. Overview of ECG-PPG multimodal fusion pipelines for cardiac monitoring applications: (a) Heart rate (HR) monitoring, illustrating ECG and PPG signal acquisition, preprocessing, peak detection, feature fusion, and final HR estimation; and (b) Heart rate variability (HRV) monitoring, showing signal acquisition, preprocessing, interval extraction, feature extraction and fusion, and HRV analysis across time-domain, frequency-domain, and nonlinear feature sets.