Analysis of heart rate variability during exercise stress testing using respiratory information

Human emotion recognition using heart rate variability analysis with spectral bands based on respiration

“…Comparing these results with the literature, HRV analysis is a suitable technique for the assessment of the balance of ANS and therefore for human emotion recognition states as stress [6], [7], [8], panic [13], [14], anxiety and depression [3] among others, even for healthy or illness people, with significant results in the spectral domain for LF, HF and LF/HF.…”

“…In order to avoid this effect, it has been proposed to center the HF band in the RF for an improved estimation of the sympathethic/parasympathetic activity [6]. The aim of this paper is to overcome this drawback in the context of HRV analysis during emotion elicitation using a HF band centered at RF, previously successfully tested in the context of stress testing [6], [7], [8]. This approach can be divided into three categories: (i) time domain analysis to obtain HRV statistical time indexes, (ii) frequency domain analysis based on nonparametric methods and (iii) same procedure as in the previous case but varying the HF spectral band based on respiration.…”

Predicting Changes in Cognitive Performance Using Heart Rate Variability

“…Note that for the SVM, kernels (RBF or linear) and their parameters C and γ in RBF kernels, C in linear kernels, whose search space was 10 −5 , 10 −4 , 10 −3 , 10 −2 , 10 −1 , 1, 10 2 , 10 3 , 10 4 , 10 5 were determined using 3-fold cross validation with a grid search to minimize mean squared error. For the RF predictor, the number of estimators, the maximum number of features, whose search space was [2,15] in RF and [2,8] in RF+PCA, were determined using 3-fold cross validation with a grid search to minimize mean squared error. Table 3 shows the performance indicators for all participants.…”

“…This indicates that it is difficult for participants in the "Not accurate" group to predict their changes in cognitive performance using HRV features because their Mean and ∆N values are quite large and their N and ∆CS I values are quite small most of the time. In addition, considering that respiration frequency, which has a negative correlation with the level of parasympathetic nerve activity, can be calculated as the peak frequency of HF [15], and that N has a negative correlation with it and that Mean has a positive correlation with it, it was assumed that the level of parasympathetic nerve activity for participants in the "Accurate" category tended to be weaker than that for participants in the "Not accurate" group.…”

“…Orsila et al showed that the level of mental stress in occupational work has a negative correlation with certain HRV features, i.e., triangular interpolation of rhythm-to-rhythm (RR) interval histogram (TINN) and RMSSD [14]. Bailon et al proposed a method for estimating mental stress during exercise using HRV and respiration features [15]. Another popular target is cognitive performance.…”