Motion Artifact Reduction in Wearable Photoplethysmography Based on Multi-Channel Sensors with Multiple Wavelengths

Figures

• 

  Figure 1. In-house-built wearable multi-channel PPG acquisition system: (a) wearable PPG hardware system consisting of a main system, sensor module (PPG and IMU), and battery; (b) the direction and wavelength (G: green; R: red; IR: infrared) of the sensor for each channel when the wearable system is worn on the wrist.

• 

  Figure 2. System diagram of the wearable multi-channel PPG acquisition system including PC application software (S/W). The solid line indicates a wired connection (USART (Universal Asynchronous Receiver/Transmitter), SPI (Serial Peripheral Interface) and I2C (Inter-Integrated Circuit)) while the dashed line indicates a wireless connection.

• 

  Figure 3. The picture of the subject being equipped with an ECG and multi-channel PPG measurement system.

• 

  Figure 4. Block diagram of signal flows for each sensor and proposed algorithm.

• 

  Figure 5. Flowchart of the proposed multi-sensor-based motion artifact reduction algorithm. (MMR is defined as Maximum to mean ratio).

• 

  Figure 6. Twelve-channel PPG signals measured while running at about 8 km/h.

• 

  Figure 7. ICA components of 12-channel PPG signals.

• 

  Figure 8. Selected independent components and motion artifact reduced PPG signals. (a) Selected ICs applying the reverse-check and pulsatile component selection algorithm, (b) reconstructed PPG signal based on the truncated SVD.

• 

  Figure 9. Example of applying the proposed method and algorithms based on best signal selection, ICA, and SVD during running at about 8 km/h.