EP4723961A1

EP4723961A1 - Method and device for evaluating human stress

Info

Publication number: EP4723961A1
Application number: EP24740206.8A
Authority: EP
Inventors: Kristjan PILT; Maie Bachmann; Ivo Fridolin; Marietta GAVRILJUK; Deniss KARAI
Original assignee: Tallinn University of Technology
Current assignee: Tallinn University of Technology
Priority date: 2023-06-12
Filing date: 2024-06-12
Publication date: 2026-04-15
Also published as: WO2024256994A1

Abstract

According to invention, changes in the forehead PPG signal, its pulsatile waveform, and SDPPG signal parameters related to arterial stiffness are used for the acute mental stress assessment.

Description

Method and device for evaluating human stress

TECHNICAL FIELD

The present invention relates to evaluating human stress, and more particularly, to a method and apparatus for evaluating human stress using photoplethysmography (PPG).

BACKGROUND ART

Social and medical problems related to mental stress are increasing globally and impact mental health and well-being of people. The acute mental stress assessment is important in everyday life to apply relaxation manoeuvres and prevent mental health issues such as anxiety and depression.

Physiological signals provide an objective assessment of mental stress. Brain activates the sympathetic nervous system, which causes an increase in heart rate, blood pressure, and the release of stress hormones, which leads to changes in blood flow and the constriction of blood vessels. In addition, arterial stiffness increases during the acute mental stress. Therefore, the arterial stiffness related parameters contribute to the model of mental stress estimation and increase its accuracy.

Numerous research studies have focused on the noninvasive detection and classification of stress levels through biological signals that are expected to express changes in the normal functioning of the body. Signals such as electroencephalogram (EEG), electrocardiogram (ECG), electromyogram (EMG), galvanic skin response (GSR), blood pressure (BP), skin temperature (SKT), photoplethysmogram (PPG) and others can extract specific information and be used as markers of stress. Since brain is the key organ, that activates many neuronal circuits when a situation is perceived as stressful, the utilization of EEG can unravel details of acute stress responses. Compared to other neuroimaging techniques, EEG modality has several advantages such as low cost, high temporal resolution and is easy to use.

The photoplethysmographic (PPG) method, which is relatively simple optical method for measuring relative blood volume changes in the artery and microvascular tissue, has been investigated for the estimation of the acute mental stress using the finger registered pulse waveform indices. The induced mental stress affects the arterial stiffness related parameters of finger PPG waveform. However, the decreased environment temperature causes vasoconstriction and drops the skin perfusion, which lowers finger PPG signal amplitude and signal to noise ratio.

Different protocols are implemented to induce stress in laboratory settings, such as the Stroop test (Stroop), cold pressor test (CPT), Trier Social Stress Test (TSST), Serial Sevens Test (SST), and others. In order to obtain the EEG signal during the subject exposure to the stress conditions, the selected protocol should be completed in a steady position to avoid any unwanted electrical input induced by the body's physical movements. Arithmetical tests can be completed with minimized movements thus SST is a suitable stressor to measure brain activity and monitor mental workload changes. Mental load is the aspect of cognitive load originating from the interaction between task and subject characteristics.

Numerical cognition is associated with a frontoparietal brain network and studies have indicated that theta band power provides evidence for the specific neural activation in this region. Additionally, changes in frontal theta band power were associated with the development of mental fatigue. While the primary stress and related mental fatigue is hidden in the normal variability of the brain activity, the cognitive task related secondary stress as a theta band power increase can be measured. This increase, whether higher or lower, can indicate the level of mental fatigue.

What is needed, therefore, is a better method for evaluating human stress.

DISCLOSURE OF INVENTION

According to the invention, the changes in the forehead PPG signal, its pulsatile waveform, and SDPPG signal parameters related to arterial stiffness are used for the acute mental stress assessment.

The invention is as claimed in attached claims 1 to 20.

BRIEF DESCRIPTION OF DRAWINGS

Embodiments of the present invention will now be described, by way of example, with reference to the accompanying drawings in which:

Fig 1. Determining the PPG DC-component (PPGDC) from the PPG signal.

Fig 2 is an Inverted PPG signal with detected rising edge 50% threshold points (top) and EKG signal (bottom) with detected R-peaks. PAT is measured as a temporal reference between the R-peak and the rising edge 50% threshold point of the PPG signal. The signals are shifted for illustrative purposes.

Fig 3 is a) filtered and averaged PPG signal, and b) the first derivative of the PPG and c) the second derivative of the PPG. Peaks 'a', 'b', 'c', 'd', and 'e' are detected from the waveform of the PPG second derivative signal.

Fig 4 is box plots of the average stress level estimates (normalized difference of combined parameters) for each locations in the city. The asterix denotes the statistically significant difference (/?<0.05) between two locations.

Fig 5 is a PPGDC dynamics during the eyes open and stress test segments for one subject.

FIG 6 is a PAT dynamics during the eyes open and stress test segments for one subject.

Fig 7 are low- and high-pass filtered PPG waveforms dring the eyes open and stress test.

Fig. 8 shows the box plots with outliers for each calculated parameter during the eyes open and stress test. The PPGb, PPGAI, b/a, c/a, d/a, e/a, AGI, and Snorm were found statistically significantly different between the two states for whole study group (p < 0.05).

Fig 9 presents the theta power values for eyes open resting and SST condition.

Fig 10 presents the normalized theta power differences for baseline and follow-up recordings.

EXAMPLES FOR CARRYING OUT THE INVENTION

Embodiment 1

Methods

Experiment setup

The forehead PPG signal was registered using TT-50 multisignal recorder (Tensiotrace, Estonia) with synchronous electrocardiographic (ECG) signal registration. The ECG signal was recorded for signal processing purposes to detect the starting points of the cardiac cycles. Both signals were digitized with a sampling rate of 500Hz. The signals were monitored online during the experiments using a Bluetooth connection.

The forehead PPG sensor was based on the AFE4490 (Texas Instruments, USA) fully integrated analogue front-end chip. The four infrared LEDs with wavelength of 865 nm (DNK1111C, Stanley Electric, Japan) were located circularly around the photodiode BPW34 (Osram, Germany). The distance between the centre of each LED and the photodiode was 8 mm. The optical components as well as the AFE4490 with peripheral components were placed in the shielded enclosure.

The PPG sensor was placed inside the cavity of the silicone holder to ensure evenly flat surface attached to the forehead. The silicone sensor holder was attached to the forehead using a wide elastic rubber band.

The blood pressure of the subject was measured using Omron M3 - HEM-7131-E (Omron, Japan) device. The cuff of the device was place to the upper arm of the dominant hand.

Subjects and experiment protocol

Citizens of a city were asked to assess the well-being of the urban environment. Therefore, the health status of the recruited subjects of this experiment was heterogenic. However, the subjects did not have any diagnosed cardiovascular (CV) diseases neither they did not take any CV system related medication. The age of the subjects had to be over 18 years. The study was approved by the Research Ethics Committee of the National Institute for Health Development with the decision no. 1064 (study no. 2351) and the research was conducted in accordance with the Helsinki Declaration.

The experiments took place in a ventilated quiet room, with white walls and natural light from the windows. The subject was first introduced for the experiment and a written informed consent was taken. Thereafter, the subject was instructed for the stress test. In this study, the modified arithmetic stress test (Series seven) was used. The subject was asked to subtract the series of sevens from the given number between 90 and 100 and the result of each subtraction had to be indicated with a pencil on a scale between 0 to 9. In case of a mistake the subject continued with the task. The subtractions started again from the given initial number in case the result was below seven.

The weight of the subject was measured and the ECG electrodes were taped to the chest. Next, the subject was asked to take a sitting position on a comfortable chair and direct the gaze to a white wall at 2 m. After 5 minutes, the blood pressure of the subject was measured and the PPG sensor was attached to the forehead. In addition, the galvanic skin response electrodes were attached to the ring and middle finger, the electroencephalograph (EEG) electrodes were placed to the forehead to both sides of the PPG sensor, and a temperature sensor was taped to the cheek. All the signals were recorded synchronously, however, in the current publication, only the forehead PPG signal analysis results are reported.

The recording of the signals was started after the attachment of all sensors and the subject was asked firstly to close their eyes for 3 minutes. Thereafter, the subject was asked to open their eyes and after 4 minutes the stress test started, which lasted for 2 minutes.

Signal processing and data analysis

PPG and ECG signals were post-processed in MATLAB (MathWorks, USA). Firstly, the 1.5 minute long segments of the eyes open and stress test were separated from the PPG and ECG signals for further processing. The R-peaks were detected from ECG signal. The PanTompkins algorithm was used for the R-peak detection.

As follows, the arterial stiffness related parameters from the PPG and SDPPG signals were calculated using previously developed and published algorithms. Briefly, the recorded raw PPG signal is inverted and filtered with window method designed (Hamming window function) low- and high-pass filters with the cut-off frequencies of 30 Hz (order of 250) and 0.5 Hz (order of 2000), respectively. Thereafter, the signal is resampled for each heart cycle in a way that the length of the selected heart cycle is 1 second. The signal is filtered with low- pass Parks-McClellan filter with an edge frequency of 6 Hz and transition band of 1 Hz (pass and stop band maximum allowable errors are 0.001), respectively. As a result, the harmonic components of each selected heart cycle is limited to 6 harmonic components. All the resampled and filtered one second long PPG waveforms are aligned according to the 50% of the ascending front and the average waveform is calculated (Fig. la). Next, the first, second, and fourth derivatives of the averaged PPG waveform are calculated. The distinct wave peaks ‘a’, ‘b’, ‘c’, ‘d’, and ‘e’ are detected from the SDPPG waveform using the fourth derivative waveform. The amplitude ratios of the SDPPG waveform peaks b/a, c/a, d/a, and e/a are calculated. The aging index (AGI) is calculated based on the SDPPG waveform amplitudes:

The amplitudes PPGb and PPGd are calculated from the PPG signal waveform based on the locations of the wave peaks ‘b’ and ‘d’ (Fig. la and c). The PPGAI is calculated as the amplitude ratio of the amplitudes PPGb and PPGd:

The slope of the ascending front of the PPG waveform (Snorm) is calculated based on the first derivative maximal point (FDPPGmax) and it is normalized with the amplitude of the PPG signal averaged waveform (APPG) as follows (see Fig. 3a and b):

The 0.002 seconds corresponds to the time between two samples of the PPG signal.

Firstly, the normal distribution of the data was tested using Anderson-Darling test. Thereafter, the statistical differences were investigated between the parameters from the eyes open and stress signal segments. The nonparametric Wilcoxon signed-rank sum test was used and the p < 0.05 was considered as statistically significant difference. The mean and standard deviation were used in case of a normally distributed dataset, otherwise, median and quartiles were calculated.

Results

The study was carried out on 42 subjects, however, the signals from 34 subjects were included in the analysis. Some of the signals were excluded due to the poor ECG electrode connections to the hairy chest, a large number of artefacts in the signals during stress test or detected arrhythmia. The general demographic, anthropometric and physiological parameters of the study group are given in Table 1. The low- and high-pass filtered PPG waveforms during the eyes open and stress test are given in Fig. 7. The heart rate during the stress test is higher as well as the differences between waveforms can be noticed. Fig. 8 shows the box plots with outliers for each calculated parameter during the eyes open and stress test. The PPGb, PPGAI, b/a, c/a, d/a, e/a, AGI, and Snorm were found statistically significantly different between the two states for whole study group (p < 0.05).

The changes in the forehead PPG waveform parameters PPGb, PPGAI, Snorm, and SDPPG signal parameters b/a, c/a, d/a, and AGI are statistically significantly different between the eyes open and the stress test states. The forehead is preferred alternative PPG signal registration location for the stress assessment purposes.

The changes in the parameters due to the induced mental stress were not in the same direction for all subjects. The arithmetic stress test may not induce the stress response equally for the subjects and depends on several factors. The physiology, as well as the previous experience of the subject to handle the stress, are some of the causes.

The forehead measurement location has skull, which is covered with a very thin soft tissue layer of the scalp. Therefore, the visible light rather reflects back from the surface of the skull and the longer wavelengths from the infrared region should penetrate to the deeper layers of the skull and through it.

Nevertheless, the measurement location for each subject was selected according to the same procedure as described in the methodology section.

The PPG waveform varies between the forehead and finger due to the measurement location anatomical differences. The arterial stiffness related waveform parameters and the parameters calculated from the finger PPG signal behave in the similar manner due to the induced stress. The induced stress increases the arterial stiffness according to the calculated parameters Snorm, PPGb, PPGAI, b/a, d/a, e/a, and AGI. From previous studies, it has been found that the aging of the subject decreases the SDPPG parameter c/a in case of the finger and forehead PPG waveform, which is expected due to the stiffening of the arteries. However, in this study, it was found that the c/a increased due to the stress, which was unexpected and needs further studies.

The results indicate that the changes in the forehead PPG and SDPPG parameters give better results compared to the finger PPG signal for the stress assessment.

Embodiment 2

Urban environment The method according to present invention was tested on 33 subjects and the physiological signals were registered in 6 different city locations in Narva (Estonia). The overview of the characteristic parameters of the test subjects is given in Table 1.1

Table 1.1

The photoplethysmographic (PPG) signal was registered from forehead using optical sensor with four LEDs in infrared region. The LEDs were positioned in PPG sensor circularly around the photodiode to the distance of 8mm. In addition, the electrocardiographic (ECG) signal and cheek temperature was registered synchronously with PPG signal. All the signals were registered with sampling rate of 500Hz.

Each subject was taken from one city location to another in random order. At each location the subject was firstly sitting with eyes closed for 3 minutes, then eyes open for 4 minutes and thereafter the stress test was conducted for 2 minutes. The arithmetic test was used as the stress test.

All the signals were postprocessed and the parameters were calculated from the signals. Firstly, the PPG signals were filtered in a way that each heart cycle contained only 6 harmonic components. Thereafter, the average waveform of PPG signal was calculated for eyes open and stress test segments. As follows, the average waveform of second derivative of PPG (SDPPG) signal was calculated and the distinct peaks were detected. The amplitude ratio of e/a from the average SDPPG signal waveform was calculated.

The R-peaks and point at 50% of PPG signal rising front was detected. The pulse arrival time (PAT), which is time delay between the R-peak and the detected point from PPG signal, for each heart cycle was calculated. The average PAT values were calculated for eyes open and stress test segments. The DC component of the PPG signal (denoted as PPGDC) was detected for each heart cycle. The average value of DC component was calculated for eyes open and stress test segments. Similarly, the average value of cheek temperature was calculated for eyes open and stress test segments.

According to the equation the stress level estimate (y) for each (z*¹¹) parameter was calculated as parameter normalized difference between eyes open and stress test segments: where the Xinitiai (also, x_baseline) and x_stress are the average values of the parameter for the eyes open and stress test segment, respectively. The average stress level estimate, S, was calculated as average over all parameters as followgs:

The average stress level estimate (normalized difference of combined parameters) was calculated for each location and each subject. The results are given in Figure 4, where the box plots for each location is given. The city locations were found to be statistically significantly different (/?<0.05) according to the stress levels. In Figure 5 are given the PPGDC component dynamics during the eyes open and stress test for one subject. Similarly, the PAT dynamics is given in Figure 6 during the eyes open and stress test for one subject.

Third embodiment

The EEG signals were recorded using Enobio 8 (Neuroelectrics Barcelona SLU) device using adhesive electrodes. The electrodes were placed on the forehead (Fpl and Fp2) according to the international 10/20 system, while two electrodes for reference and ground were placed behind the right ear. Raw EEG signals were recorded at a sampling frequency of 500 Hz.

The recorded EEG signals were digitally filtered at the cut-off frequencies of 3 Hz and 47 Hz. The signals were visually inspected and signal segments with artefacts were removed. Therefore, the further analysis was performed on 110 second EO resting and SST EEG signal segments from 39 subjects, where EO resting and SST segments were obtained as the last 110 seconds of the corresponding segment. MATLAB (The MathWorks, Inc.) software was used for EEG signal processing. First, each 110 sec EEG signal segment was cut into 10 sec epochs. Power spectral density (PSD) was estimated by means of the Welch averaged periodogram method. Epochs were divided into overlapping sections (50%), with a length of 1024 points. Thereafter, theta band power was calculated at 4-7 Hz for each epoch. After that, epoch-based median theta power was obtained for resting and SST segments. The normalized theta power difference of rest and SST conditions was calculated as:

The normalized theta power difference represents the difference between the theta band power in the EO cognitive arithmetic task condition (0 PSST) and the theta band power in the EO resting condition (O PRest), where the normalization is performed relative to their average.

The paired sample Wilcoxon signed rank test was applied for the statistical comparison of the normalized theta power difference between baseline and follow-up recordings. The confidence level of 0.05 was considered statistically significant to the Wilcoxon p values.

Fig 9 presents the theta power values for eyes open resting and SST condition. There is a statistically significant increase in theta power in SST condition (p=0.0011). The results are in agreement with the review performed by Chikhi et al. concluding that increased cognitive load is often indicated in increased frontal theta band power.

Due to the high theta power variability in the group (interindividual differences), the normalized theta power difference between rest and SST was compared to analyze the changes for baseline and follow-up recordings at group level.

Fig 10 presents the normalized theta power differences for baseline and follow-up recordings. The positive median values for baseline and follow up condition indicate the increased theta power in STT in both conditions. Still, the difference is statistically significantly higher for baseline recordings compared to follow-up recordings indicating higher brain activity on calculations during baseline, while at follow-up the cognitive workload on calculations is lower. As the location where the test was performed was the same for both conditions, the difference cannot be explained by the recording environment. It is presumed that during follow-up the subjects had higher fatigue levels, so the amount of available cognitive resources for STT was lower.

As the EEG reflects the brain state which in turn is sensitive to recording conditions, it is suggested to perform the EEG studies always at the same time of day, while the morning is preferred as the brain is not yet loaded with different everyday assignments. On the other hand, the sensitivity to recording conditions could also be put to work as an advantage - the theta band power together with a cognitive test could be used to evaluate the stress caused by specific environments etc.

Current study has some limitations to consider. As the group’s overall variability of the normalized theta power difference was high, considering the age could decrease the overall variability. It is presumed that SST has age limitations and there is an age group for whom this method works better than others. Present study did not consider males and females separately, while it has been found that the EEG power at different conditions varies for males and females.

Acknowledgements

This work was supported partly by the Estonian Ministry of Research and Education and European Regional Development Fund (grant 2014-2020.4.01.20-0289) and by the Estonian Research Council grant (PSG819).

References

1. Steel, Z., Marnane, C., Iranpour, C., Chey, T., Jackson, J. W., Patel, V., Silove, D.: The global prevalence of common mental disorders: a systematic review and meta-analysis 1980- 2013. Int J Epidemiol 43(2), 476-493 (2014).

2. World Health Organization: WHO Mental health atlas 2014. WHO, Geneva, 69 p (2015).

3. Hamdani, S.U., Zill-e-Huma, Zafar, S.W., Suleman N., Um-Ul-Baneen, Waqas A., Rahman A.: Effectiveness of relaxation techniques ‘as an active ingredient of psychological interventions’ to reduce distress, anxiety and depression in adolescents: a systematic review and meta-analysis. Int J Ment Health Syst 16(1), 31 (2022).

4. Arza, A., Garzon-Rey, J. M., Lazaro, J., Gil, E., Lopez-Anton, R., de la Camara, C., Laguna, P., Bailon, R., Aguilo, J.: Measuring acute stress response through physiological signals: towards a quantitative assessment of stress. Med Biol Eng Comput 57(1), 271-287 (2019).

5. Vlachopoulos, C., Kosmopoulou, F., Alexopoulos, N., loakeimidis, N., Siasos, G., Stefanadis, C.: Acute mental stress has a prolonged unfavorable effect on arterial stiffness and wave reflections. Psychosom Med 68(2), 231-237 (2006).

6. Rinkevicius, M., Kontaxis, S., Gil, E., Bailon, R., Lazaro, J., Laguna, P., Marozas, V.: Photoplethysmogram Signal Morphology -Based Stress Assessment. In: 2019 Computing in Cardiology (CinC), pp. Page 1-Page 4. IEEE, Singapore (2019).

7. Pilt, K., Karai, D., Bachmann, M., Gavriljuk, M., Allik, A., Uudeberg, T., Tarnov, K., Fridolin, I.: Influence of mental stress on the arterial stiffness parameters detected from photoplethysmographic signal waveform. In: 2022 18th Biennial Baltic Electronics Conference (BEC), pp. 1-4. IEEE, Tallinn, Estonia (2022)

8. Fine, J., Branan, K. L., Rodriguez, A. J., Boonya-Ananta, T., Ajmal, Ramella-Roman, J. C., McShane, M. J., Cote, G. L.: Sources of Inaccuracy in Photoplethysmography for Continuous Cardiovascular Monitoring. Biosensors (Basel). 11(4), 126 (2021).

9. Allen J.: Photoplethysmography and its application in clinical physiological measurement. Physiol Meas 28(3), Rl-39 (2007).

10. Mannheimer, P. D., O', N. M., Konecny, E.: The influence of larger subcutaneous blood vessels on pulse oximetry. J Clin Monit Comput 18(3), 179-88 (2004).

11. Pilt, K., Meigas, K., Temitski, K., Viigimaa, M.: Second derivative analysis of forehead photoplethysmographic signal in healthy volunteers and diabetes patients. In: Long, M. (eds) World Congress on Medical Physics and Biomedical Engineering 2012, IFMBE Proceedings, vol. 39, pp. 410-413. Springer, Heidelber (2013).

12. Pilt, K., Meigas, K., Ferenets, R., Temitski, K., Viigimaa, M.: Photoplethysmographic signal waveform index for detection of increased arterial stiffness. Physiol Meas 35(10), 2027-2036 (2014).

13. Pilt, K., Meigas, K., Kbbts, K., Viigimaa, M.: Photoplethysmographic signal rising front analysis for the discrimination of subjects with increased arterial ageing. Proc Est Acad Sci 63(3), 309-314 (2014). 14. Takazawa, K., Tanaka, N., Fujita, M., Matsuoka, O., Saiki, T., Aikawa, M., Tamura, S., Ibukiyama, C.: Assessment of vasoactive agents and vas-cular aging by the second derivative of photoplethysmogram waveform. Hypertension 32(2), 365-370 (1998).

15. Thoits, P. A.: Stress and health: major findings and policy implications. J Health Soc Behav, 51, 41-53 (2010).

16. Tomiyama, A. J.: Stress and Obesity. Annu Rev Psychol, 70, 703-718 (2019).

17. Cohen, S., Kamarck, T., Mermelstein, R.: A global measure of perceived stress. J Health Soc Behav. 24(4), 385-396 (1983).

18. Bjbmtorp, P.: Do stress reactions cause abdominal obesity and comorbidities?, Obesity Reviews, 2(2), 73-86 (2001).

19. Ladakis, I., Chouvarda, I.: Overview of biosignal analysis methods for the assessment of stress. Emerging Science Journal, 5(2), 233-244 (2021).

20. de Kloet, E.R., Joels, M., Holsboer, F.: Stress and the brain: from adaptation to disease. Nat Rev Neurosci. 6(6), 463-75 (2005).

21. Aspiotis, V., Miltiadous, A., Kalafatakis, K., Tzimourta, K.D., Giannakeas, N., Tsipouras, M.G., Peschos, D., Glavas, E., Tzallas, A.T.: Assessing Electroencephalography as a Stress Indicator: A VR High-Altitude Scenario Monitored through EEG and ECG. Sensors (Basel). 22(15), 5792 (2022).

22. Spiller, M., Walter, C., Rosenstiel, W. et al : EEG-based prediction of cognitive workload induced by arithmetic: a step towards online adaptation in numerical learning. ZDM Mathematics Education 48, 267-278 (2016).

23. Trejo, L., Kubitz, K., Rosipal, R., Kochavi, R., Montgomery, L.: EEG-Based Estimation and Classification of Mental Fatigue. Psychology, 6, 572-589 (2015).

24. Trejo, L., Kubitz, K., Rosipal, R., Kochavi, R. & Montgomery, L.: EEG-Based Estimation and Classification of Mental Fatigue. Psychology, 6, 572-589 (2015).

25. Chikhi, S., Matton, N., Blanchet, S.: EEG power spectral measures of cognitive workload: A meta-analysis. Psychophysiology. 59(6), 14009 (2022). 26. Tran, Y., Craig, A., Craig, R., Chai, R., Nguyen, H. The influence of mental fatigue on brain activity: Evidence from a systematic review with meta-analyses. Psychophysiology. 257(5):el3554 (2020). doi: 10.111 l/psyp.13554.

27. Cantillo-Negrete, J., Carino-Escobar, R.I., Carrillo-Mora, P., Flores-Rodriguez, T.B., Elias-Vinas, D., Gutierrez -Martinez, J.: Gender Differences in Quantitative Electroencephalogram During a Simple Hand Movement Task in Young Adults. Rev Invest Clin. 68(5), 245-255 (2016).

Claims

1. A method of evaluating an individual’s stress using photoplethysmography PPG, comprising recording said PPG signal from the forehead of the individual.

2. Method according to claim 1, characterized in that the PPG signal has an infrared wavelength.

3. Method according to claim 2, characterized in that that a distance between at least one light source and a detector device is from 6 to 12 mm.

4. Method according to claim 3, characterized in that for stress assessment, a mean value of the DC component of the PPG signal (PPGDC) is recorded over time interval ranging from 10 seconds to 5 minutes.

5. Method according to claim 4, characterized in that the mean value of the PPGDC is determined during an initial state, denoted as x initial, and during a stress-induced state, denoted as x_stress, using a stress test, e.g., an arithmetic test.

6. Method according to claim 5, characterized in that a stress level estimation is computed based on the difference between the PPGDC during the initial state and the stress-induced state.

7. Method according to claim 6, characterized in that the stress level estimation, y(i), for an individual is calculated using the equation:

8. Method according to claim 4, characterized in that an electrocardiographic EKG signal and cheek temperature are measured from the individual.

9. Method according to claim 8, characterized in that the R-peak is detected from the

EKG signal for each cardiac cycle, and a point corresponding to the arrival of the pulse wave is detected from the forehead signal.

10. Method according to claim 9, characterized in that for each cardiac cycle, the point corresponding to the rising edge of the forehead PPG signal, which represents the arrival of the pulse wave is detected.

11. Method according to claim 10, characterized in that the pulse arrival time (PAT), which is the time delay between the R-peak and the detected pulse arrival point in the forehead PPG signal, e.g., between the rising edge's 50% threshold points, is determined for each cardiac cycle.

12. Method according to claim 3, characterized in that the alternating component is extracted from the forehead PPG signal, and parameters characterizing arterial stiffness are calculated.

13. Method according to claim 12, characterized in that the PPG alternating component signal is filtered such that each cardiac cycle contains 6 to 10 harmonic components.

14. Method according to claim 13, characterized in that the second derivative signal of the PPG alternating component is computed, and wave peaks and their relationships are detected.

15. Method according to claim 14, characterized in that the amplitude ratio of the second derivative PPG signal (e/a), pulsewave arrival time (PAT), and cheek temperature are measured during the baseline state and during the stress-induced state using a stress test such as an arithmetic test.

16. Method according to claim 15, characterized in that a stress level estimation is computed for each physiological parameter based on the difference between the initial and stress-induced states.

17. Method according to claim 16, characterized in that the stress level, y, for each i-th parameter (i = 1..N) is determined using the equation: where xinitiai(i) is the mean value or other statistically characteristic parameter determined during the baseline state for the i-th parameter, and x_stress(i) is the central value or other statistically characteristic parameter determined during the stress test for the i-th parameter.

18. Method according to claim 17, characterized in that the individual's stress level estimation is computed as a combination of the stress level estimations y(i) calculated for each parameter.

19. Method according to claim 18, the individual's stress level, S, is computed as the arithmetic mean of the stress level estimations calculated for each parameter:

20. A portable device, comprising a set of optical sensors to be attached to the forehead of an individual and a processor device, adapted to receive signals from said set of sensors, said processor device adapted to carry out methods according to any of the claims 1 to 19.