Autonomic and neural responses to varying transcutaneous cervical electrical stimulation parameters

Fig. 1. A Descriptions of testing sessions. For each of four visits, one stimulation parameter was varied, shown in red, while all others remained constant. In subsequent visits, the next parameters were chosen based on the largest RMSSD responses in previous testing sessions. For each condition, a 12-min block consisted of baseline recording for 4 min, stimulation for 4 min, and a post-stimulation recording for 4 min. Based on the number of options for each parameter, participants went through 3 to 4 of these blocks per visit. In visit 4, frequency was varied like in visit 1; the frequency that resulted in the largest RMSSD response was then used to test stimulation duration in a 20-min stimulation block. B Physiological monitoring. Signals for each participant were recorded by a six-lead electrocardiogram (in red, wires attached to four foam adhesive electrodes placed at each shoulder and each ankle) and noninvasive blood pressure (in blue, small inflatable cuff on middle phalanx of middle finger). A dry electroencephalography system was placed on the participant’s head, while eye tracking glasses were used to record pupil diameter and gaze location. C Bilateral stimulation montage.…

Fig. 2. Normalized post-stimulation responses to varying frequencies, visit 1. Boxplots show normalized post-stimulation responses for A HR, B MAP, C RMSSD, D SDNN, E PDL, F delta, G theta, H alpha, and I beta band power. For each boxplot, the red horizontal line denotes the median, while the box is determined by the 25th to 75th percentiles. Whiskers extended to the maximum and minimum values that are less than 1.5 times the interquartile range away from the bottom or top of the box; outliers are beyond this limit. Individual responses are displayed by the colored points and connected by dotted lines

Fig. 3. Personalizing stimulation frequency, visit 1. A Histogram of preferred frequencies based on the largest RMSSD increase. B This favored frequency occurred at random during the testing visit, though it most often was the first or last stimulation block. C Shown are the average RMSSD responses at each frequency (colored lines) compared to the average RMSSD response to the favored frequency for each individual (black line). D Histogram of preferred frequencies based on largest alpha band power increase. E This favored frequency occurred at random during the testing visit, though it most often was the third or last stimulation block. F Shown are the average alpha band power responses at each frequency (colored lines) compared to the average alpha band power response to the favored frequency for each individual (black line)

Fig. 4. Normalized post-stimulation responses to varying intensities, visit 2. Boxplots show normalized post-stimulation responses for A HR, B MAP, C RMSSD, D SDNN, E PDL, F delta, G theta, H alpha, and I beta band power. For each boxplot, the red horizontal line denotes the median, while the box is determined by the 25th to 75th percentiles. Whiskers extended to the maximum and minimum values that are less than 1.5 times the interquartile range away from the bottom or top of the box; outliers are beyond this limit. Individual responses are displayed by the colored points and connected by dotted lines

Fig. 5. Comparing stimulation intensity. A Histogram of stimulation intensities with the largest RMSSD increase. B This RMSSD response occurred at random during the testing visit. C Shown are the average RMSSD responses at each intensity (colored lines) at each individual’s favored frequency. D Histogram of stimulation intensities with the largest alpha band power increase. E This alpha band power response occurred at random during the testing visit. F Shown are the average alpha band power responses at each intensity (colored lines) at each individual’s favored frequency

Fig. 6. Normalized post-stimulation responses to varying montages, visit 3. Boxplots show normalized post-stimulation responses for A HR, B MAP, C RMSSD, D SDNN, E PDL, F delta, G theta, H alpha, and I beta band power. Visit 1 (V1) and Visit 2 (V2) have matched parameters with Visit 3’s Bilateral stimulation. For each boxplot, the red horizontal line denotes the median, while the box is determined by the 25th to 75th percentiles. Whiskers extended to the maximum and minimum values that are less than 1.5 times the interquartile range away from the bottom or top of the box; outliers are beyond this limit. Individual responses are displayed by the colored points and connected by dotted lines

Fig. 7. Comparing electrode montages. A Histogram of montage laterality with the largest RMSSD increase. B This RMSSD response occurred at random during the testing visit. C Shown are the average RMSSD responses at each montage (colored lines) at each individual’s favored frequency and high stimulation intensity. D Histogram of montage laterality with the largest alpha band power increase. E This alpha band power response occurred at random during the testing visit. F Shown are the average alpha band power responses at each montage (colored lines) at each individual’s favored frequency and high stimulation intensity

Fig. 8. Normalized post-stimulation responses to varying frequencies, comparing visits 1 and 4. Boxplots show normalized post-stimulation responses for A HR, B MAP, C RMSSD, D SDNN, E PDL, F delta, G theta, H alpha, and I beta band power. For each boxplot, the red horizontal line denotes the median, while the box is determined by the 25th to 75th percentiles. Whiskers extended to the maximum and minimum values that are less than 1.5 times the interquartile range away from the bottom or top of the box; outliers are beyond this limit. Individual responses are displayed by the colored points and connected by dotted lines

Fig. 9. Personalizing stimulation frequency, visit 4. A Histogram of preferred frequencies based on the largest RMSSD increase. B This favored frequency occurred at random during the testing visit, though it most often was the third stimulation block. C Shown are the average RMSSD responses at each frequency (colored lines) compared to the average RMSSD response to the favored frequency for each individual (black line), with high stimulation intensity and right-sided stimulation. D Histogram of preferred frequencies based on largest alpha band power increase. E This favored frequency occurred at random during the testing visit, though it most often was the third or last stimulation block. F Shown are the average alpha band power responses at each frequency (colored lines) compared to the average alpha band power response to the favored frequency for each individual (black line), with high stimulation intensity and right-sided stimulation

Fig. 10. Normalized post-stimulation responses to varying duration, visit 4. Boxplots show normalized post-stimulation responses for A HR, B MAP, C RMSSD, D SDNN, E delta, F theta, G alpha, and H beta band power. For each boxplot, the red horizontal line denotes the median, while the box is determined by the 25th to 75th percentiles. Whiskers extended to the maximum and minimum values that are less than 1.5 times the interquartile range away from the bottom or top of the box; outliers are beyond this limit. Individual responses are displayed by the colored points and connected by dotted lines

Fig. 11. Validity of PPG-derived RR-intervals and PRV response to stimulation. A Shown is a scatter plot of all extracted RR interval values from PPG vs. those extracted from ECG in non-stimulation periods, along with a linear regression trendline (black, y = 1.0639x-0.0539, R.2 = 0.7439). The unity line (y = x) is shown in red. B Using RR intervals extracted from PPG, RMSSD as a pulse rate variability measure can be calculated during the 20-min stimulation period. The average over all participants is shown in black, while other colors denote the average of all participants with the same preferred frequency, determined by the largest RMSSD (from RR intervals extracted from ECG) from the 4-min stimulation in visit 4