Have a personal or library account? Click to login
Storing information electrically in human skin Cover

Storing information electrically in human skin

Journal of Electrical Bioimpedance
Volume 12 (2021): Issue 1 (January 2021)
By: Oliver Pabst,  Øystein Magnus Sørebø,  Karoline Sjøen Andersen,  Erlend Lemva Ousdal,  Sean William Bråthen,  Badi Ur Rehman,  Haiatullah Gholami,  Zhijian Zhou,  Koki Takahashi,  Diriba Tasfaye Dumesso,  Mellie Merete Livingston,  Wesley Julian Lodewijk,  Stian Sæther,  Alireza Eskandari Turk and  Peter Louis Uller  
Open Access
|Nov 2021

Figures & Tables

Fig. 1

Schematic of the used instrumentation (same schematic as presented in [9] under Creative Commons Attribution 4.0 International License), electrode placement and the two chosen numbers and the corresponding states (0,1,2,3) under each electrode. The used three electrode setup [13] with CC as the current carrying electrode and Ref as the reference electrode enables monopolar recordings under the measurement electrodes M1, M2, and M3. For the first 6 out of 17 subjects, the three measurement electrodes were placed horizontally along the forehead as shown in variant A. For the remaining 11 out of 17 subjects, the measurement electrodes were placed vertically along the forehead above the nose as shown in variant B.
Schematic of the used instrumentation (same schematic as presented in [9] under Creative Commons Attribution 4.0 International License), electrode placement and the two chosen numbers and the corresponding states (0,1,2,3) under each electrode. The used three electrode setup [13] with CC as the current carrying electrode and Ref as the reference electrode enables monopolar recordings under the measurement electrodes M1, M2, and M3. For the first 6 out of 17 subjects, the three measurement electrodes were placed horizontally along the forehead as shown in variant A. For the remaining 11 out of 17 subjects, the measurement electrodes were placed vertically along the forehead above the nose as shown in variant B.

Fig. 2

Results of one whole recording of test subject “A” before exercise. Parameterization and durations are visualized. The initial conductance was measured for around one minute in step 1. In step 2, DC voltage pulses were applied to change the conductance of the skin (non-linear measurements). The procedure was different for the two different numbers (302 and 123); this example shows the procedure for number 302. An example measurement for 123 can be seen in Fig. 3. In step 3, the AC conductance was measured for 20 minutes. The start of each measurement was done manually within the software and thus the durations t1, t2, and t3 between measurements varied slightly among the test sessions. The initial conductance, here defined as the last conductance value within step 1 is labeled as Gin. Within step 2, iin and iend are the current values at the beginning and the end of each DC voltage pulse series. For the further analysis, the conductance values at specific times (3, 5, 10, 15, and 20 minutes) related to the time of the last DC pulse (which is different for each electrode) were used. In the plot to the right (step 3), only the conductance values at minute five, G5min, are indicated.
Results of one whole recording of test subject “A” before exercise. Parameterization and durations are visualized. The initial conductance was measured for around one minute in step 1. In step 2, DC voltage pulses were applied to change the conductance of the skin (non-linear measurements). The procedure was different for the two different numbers (302 and 123); this example shows the procedure for number 302. An example measurement for 123 can be seen in Fig. 3. In step 3, the AC conductance was measured for 20 minutes. The start of each measurement was done manually within the software and thus the durations t1, t2, and t3 between measurements varied slightly among the test sessions. The initial conductance, here defined as the last conductance value within step 1 is labeled as Gin. Within step 2, iin and iend are the current values at the beginning and the end of each DC voltage pulse series. For the further analysis, the conductance values at specific times (3, 5, 10, 15, and 20 minutes) related to the time of the last DC pulse (which is different for each electrode) were used. In the plot to the right (step 3), only the conductance values at minute five, G5min, are indicated.

Fig. 3

Recordings of test subject “B” before (top plots, number 123 in this case) and after (bottom plots, number 302 in this case) physical exercise shown for the skin under all three measurement electrodes (M1, M2, and M3). DC voltage pulses were applied (middle) and small signal conductance was measured before (left) and after (right) for around 1 minute, and 20 minutes, respectively. The drop of the conductance to zero under M3 in the 20 minutes conductance measurement is an artifact.
Recordings of test subject “B” before (top plots, number 123 in this case) and after (bottom plots, number 302 in this case) physical exercise shown for the skin under all three measurement electrodes (M1, M2, and M3). DC voltage pulses were applied (middle) and small signal conductance was measured before (left) and after (right) for around 1 minute, and 20 minutes, respectively. The drop of the conductance to zero under M3 in the 20 minutes conductance measurement is an artifact.

Fig. 4

Boxplots based on all measurements with number 123 (left, N=19) and number 302 (right, N=13). a) Boxplots of iRatio. In case of number 123, no DC voltage pulses were applied to M1, three positive pulses were applied to M2 and 5 to M3. In case of number 302, 5 positive DC pulses were applied to M1, 3 negative pulses were applied to M2 and 3 positive pulses were applied to M3. b) Boxplots of Gratio over time for the three different M electrodes. States 1, 2, and 3 (left plot, N=19) were stored in the skin under the electrodes M1, M2, and M3 respectively. Friedman test and pairwise comparison (Tukey test) show that there are significant differences between state 2 and state 1 and also between states 3 and 1 but not between states 2 and 3 at all time points (see also table 2 in the supplemental information). States 3, 0, and 2 (right plot, N=13) were stored in the skin under the electrodes M1, M2, and M3 respectively. Similar as for number 123, Friedman test and pairwise comparison (Tukey test) show that there are significant differences between state 2 and state 0 and also between states 3 and 0 but not between states 2 and 3 at all time points. There is no data for M2 (state 0) at minute 3 since 190 seconds of positive DC voltage pulses followed before the small signal conductance was conducted (see Fig. 2).
Boxplots based on all measurements with number 123 (left, N=19) and number 302 (right, N=13). a) Boxplots of iRatio. In case of number 123, no DC voltage pulses were applied to M1, three positive pulses were applied to M2 and 5 to M3. In case of number 302, 5 positive DC pulses were applied to M1, 3 negative pulses were applied to M2 and 3 positive pulses were applied to M3. b) Boxplots of Gratio over time for the three different M electrodes. States 1, 2, and 3 (left plot, N=19) were stored in the skin under the electrodes M1, M2, and M3 respectively. Friedman test and pairwise comparison (Tukey test) show that there are significant differences between state 2 and state 1 and also between states 3 and 1 but not between states 2 and 3 at all time points (see also table 2 in the supplemental information). States 3, 0, and 2 (right plot, N=13) were stored in the skin under the electrodes M1, M2, and M3 respectively. Similar as for number 123, Friedman test and pairwise comparison (Tukey test) show that there are significant differences between state 2 and state 0 and also between states 3 and 0 but not between states 2 and 3 at all time points. There is no data for M2 (state 0) at minute 3 since 190 seconds of positive DC voltage pulses followed before the small signal conductance was conducted (see Fig. 2).

Fig. 5

Initial conductance values, Gin, under each electrode (M1, M2, and M3) shown for all 17 test subjects before and after exercise. The plot in the lower right corner is a zoomed in version of the plot to the left.
Initial conductance values, Gin, under each electrode (M1, M2, and M3) shown for all 17 test subjects before and after exercise. The plot in the lower right corner is a zoomed in version of the plot to the left.

Fig. 6

Recordings of test subject “C” before (top plots) and after physical exercise (bottom plots) shown for the skin under all three measurement electrodes (M1, M2, and M3). For this subject, both measurements were done with number 123. This subject was electrodermal active at the forehead.
Recordings of test subject “C” before (top plots) and after physical exercise (bottom plots) shown for the skin under all three measurement electrodes (M1, M2, and M3). For this subject, both measurements were done with number 123. This subject was electrodermal active at the forehead.

Possible combinations of applied numbers before and after exercise and obtained number of test subjects (N) per combination_

Statistical evaluation of Gratio(t) at time points 3, 5, 10, 15 and 20 minutes over all recordings with number 123 (top, N=19) and number 302 (bottom, N=13)_ Median and 25% and 75% percentiles of Gratio(t) are presented, followed by the results from Friedman test and the results from the pairwise comparison (Tukey test)_ Within group comparisons of the measurements with number 123 show that there are significant differences (see P-values) between states 3 and 1 and also between states 2 and 1 but not within states 2 and 3 (P-values > 0_05)_ This is valid for all picked times_ Within group comparisons of the measurements with number 302 show that there are significant differences between states 3 and 0 and also between states 2 and 0 but not within states 2 and 3, which is similar to the measurements with number 123_

DOI: https://doi.org/10.2478/joeb-2021-0010 | Journal eISSN: 1891-5469
Journal RSS Feed
Language: English
Page range: 73 - 81
Submitted on: Jun 21, 2021
Published on: Nov 29, 2021
Published by: University of Oslo
In partnership with: Paradigm Publishing Services
Publication frequency: 1 issue per year
Keywords:
Bioimpedance,
human skin,
memristor,
information storage,
non-linear electrical measurements
Related subjects:
Engineering,
Bioengineering and biomedical engineering,
Biomedical electronics,
Life sciences,
Biophysics,
Medicine,
Biomedical engineering,
Physics,
Spectroscopy and metrology

© 2021 Oliver Pabst, Øystein Magnus Sørebø, Karoline Sjøen Andersen, Erlend Lemva Ousdal, Sean William Bråthen, Badi Ur Rehman, Haiatullah Gholami, Zhijian Zhou, Koki Takahashi, Diriba Tasfaye Dumesso, Mellie Merete Livingston, Wesley Julian Lodewijk, Stian Sæther, Alireza Eskandari Turk, Peter Louis Uller, published by University of Oslo
This work is licensed under the Creative Commons Attribution 4.0 License.

Previous articleVolume 12 (2021): Issue 1 (January 2021)Next article