Estimating psychophysiological loads by repeated temperature steps on humans using a state–space model

Figures

• 

  Fig 1. Schematic representation of the detailed study protocol.

• 

  Fig 2. Subjective index and skin temperature. Results of (a) subjective physical fatigue and (b) skin temperature modelling. The left charts show the superimposed model estimates and mean values of the actual measurements for 28 participants. The right charts show the Load term, which is the accumulated load. In the left charts, empty markers represent the posterior average of the modelling, and filled markers represent the average of the actual measured data for 28 participants. The light band represents the 95% credible interval, and the dark band represents the 50% credible interval for each condition. The light orange background indicates room B. The x-axis shows the time points, and the y-axis shows the measured and model-estimated values (posterior mean) of each index.

• 

  Fig 3. Heart rate variability. Results of (a) HF (high-frequency component of heart rate variability), (b) LF/HF (the ratio between low- frequency component and high-frequency component of heart rate variability), and (c) HR (heart rate) modelling. The left charts show the superimposed model estimates and mean values of the actual measurements for 28 participants. The right charts show the Load term, which is the accumulated load. In the left charts, empty markers represent the posterior average of the modelling, and filled markers represent the average of the actual measured data for 28 participants. The light band represents the 95% credible interval, and the dark band represents the 50% credible interval for each condition. The light orange background indicates room B. The x-axis shows the time points, and the y-axis shows the measured and model-estimated values (posterior mean) of each index.

• 

  Fig 4. Prediction of subjective index and skin temperature. Results of the (a) subjective physical fatigue and (b) skin temperature predictions. Model estimates to the left of the vertical black line are based on actual measurements at the corresponding time points, and the models after the fourth set on the right side of the black vertical line are estimated using the measured values up to the third set. The colour band represents the 50% credible interval. The light orange background indicates room B. The x-axis shows the time points, and the y-axis shows the measured value, estimated model, and predicted model of each index. Filled markers represent the measured values, and empty markers represent the estimated models.

• 

  Fig 5. Prediction of heart rate variability. Results of (a) HF (high-frequency component of heart rate variability), (b) LF/HF (the ratio between low- frequency component and high-frequency component of heart rate variability), and (c) HR (heart rate) modelling. Model estimates to the left of the vertical black line are based on actual measurements at the corresponding time points, and models after the fourth set on the right side of the black vertical line are estimated using the measured values up to the third set. The colour band represents the 50% credible interval. The light orange background indicates Room B. The x-axis shows the time points, and the y-axis shows the measured value, calculated value, and predicted value of each index. Filled markers represent the measured values, and empty markers represent the estimated models.