Evaluating the Potential of an Oral-Based Bioguard to Estimate Heart Rate Using Photoplethysmography

Figures

• 

  Figure 1. Point of placement for the front and back photoplethysmography sensors. The front sensor was located at the superior alveolar process above the right maxillary lateral incisor to observe blood flow from the middle superior alveolar artery. The back sensor was located at the palate behind the left upper first molar to measure blood flow at the arterioles of the greater palatine artery.

• 

  Figure 2. An EVA-encased mouthguard was situated on top of the cast with two MAX30102 sensors embedded. A cable was used to connect it to the MAX30102ACCEVKIT Evaluation System. Green electric tape was used to keep the wires in place during the forming process, while a yellow pH-sensitive comparator paper strip was positioned along the wires to detect potential water ingress.

• 

  Figure 3. A dental cast with red tape marking the area beyond the gumline likely to contain a high density of arterioles.

• 

  Figure 4. Model of light propagation in the mouthguard. (A) Baseline cast without tape. Light is expected to scatter both in the EVA and in the cast material. (B) Taped cast. The red tape should change the reflection of the red and infrared light in comparison to the bare cast. (C) The PPG positioned closer to the teeth area should have most of the light reflected by the cast material. (D) The PPG positioned in between the teeth area and the gum should reflect light affected by an area of only cast material and an area covered by the red tape. (E) The PPG is positioned completely in the gum area and most of the light reflection should be affected by the taped cast.

• 

  Figure 5. Flowchart of signal data processing.

• 

  Figure 6. Boxplots of the error of heart rate calculation from the PPG without wavelet denoising separated by error source predictors. The subjects’ boxplot is displayed to expose the difference in the manufactured mouthguards that was assessed using the gum reflection metric. The central lines of the boxplots are the median, the whiskers are the upper and lower quartiles and the outliers are points 1.5 times above the upper and lower quartiles.

• 

  Figure 7. Scatter plot of the error of heart rate calculation from the PPG without wavelet denoising separated by gum reflection metric. Each participant had four gum reflection metrics that consisted of front red, front infrared, back red and back infrared. All data are shown in this plot.

• 

  Figure 8. Boxplots of the error of the heart rate calculation from the PPG separated by error source predictors. The subjects’ boxplot is displayed to expose the difference in the manufactured mouthguards that was assessed using the gum reflection metric. The central lines of the boxplots are the median, the whiskers are the upper and lower quartiles and the outliers are points 1.5 times above the upper and lower quartiles.

• 

  Figure 9. Scatter plot of the error of the heart rate calculation from the PPG separated by the gum reflection metric. Each participant had four gum reflection metrics that consisted of front red, front infrared, back red and back infrared. All the data are shown in this plot.

• 

  Figure 10. Example of raw (top figures) and filtered (bottom figures) PPG signals at a light intensity of 5.08 mA. The left side shows the data for the infrared (IR) and the right side for the red LED.