Mental Stress Assessment Using Ultra Short Term HRV Analysis Based on Non-Linear Method

Abstract

Mental stress is on the rise as one of the major health problems in modern society. It is important to detect and manage mental stress to prevent various diseases caused by stress and to maintain a healthy life. The purpose of this paper is to present new heart rate variability (HRV) features based on empirical mode decomposition and to detect acute mental stress through short-term HRV (5 min) and ultra-short-term HRV (under 5 min) analysis. HRV signals were acquired from 74 young police officers using acute stressors, including the Trier Social Stress Test and horror movie viewing, and a total of 26 features, including the proposed IMF energy features and general HRV features, were extracted. A support vector machine (SVM) classification model is used to classify the stress and non-stress states through leave-one-subject-out cross-validation. The classification accuracies of short-term HRV and ultra-short-term HRV analysis are 86.5% and 90.5%, respectively. In the results of ultra-short-term HRV analysis using various time lengths, we suggest the optimal duration to detect mental stress, which can be applied to wearable devices or healthcare systems.

Feasibility of Ultra-Short-Term Analysis of Heart Rate and Systolic Arterial Pressure Variability at Rest and during Stress via Time-Domain and Entropy-Based Measures

“…However, the results of correlation analysis have evidenced that the agreement between UST and ST distributions is overall very good (i.e., above threshold) only for N ≥ 120 for RR and for N ≥ 180 for SAP, except for non-parametric SE that already exhibits a severe decrease of from 4 min length recordings. The results of RR analyses are in agreement with some previous studies employing other non-linear measures for the analysis of predictability (e.g., Shannon Entropy [ 29 , 33 ]), dynamics (e.g., Permutation Entropy [ 79 ]), and complexity (e.g., Approximate Entropy and Sample Entropy [ 29 , 33 , 79 ]), overall reporting that recordings of at least 2–3 min are necessary in order to have good consistency with respect to ST standard. Our results complement these findings, supporting the hypothesis that the variation in cardiovascular dynamics and the complexity produced by a physiological state change can be properly assessed using even shorter recordings (60 samples), but at the cost of a lower correlation with ST reference.…”

Influences of Intense Physical Effort on the Activity of the Autonomous Nervous System and Stress, as Measured with Photoplethysmography

“…The use of algorithms, artificial intelligence (AI), and frequency domain analysis of HRV embedded in apps and devices provides easy interpretations of these data that can be used to improve the quality of life and health of patients and users [ 28 ]. Max Pulse ® , like other modern devices, is based on the use of these technological advances for the interpretation of results and the automatic use of data, facilitating the investigations of researchers in the assessment and quantification of stress [ 29 , 30 , 31 , 32 ]. The novelty of this device is that it facilitates the interpretation of the data associated with the influence of HRV on the ANS, both in absolute values and when comparing between two different situations, in our case before and after intense exercise.…”

Machine Learning models for the Cognitive Stress Detection Using Heart Rate Variability Signals

“…The LF/HF, however, did not consistently increase during the cognitive load task. Although many HRV-based machine learning models have used this metric as a predictor of certain behavioral outcomes [17,50,51], there is a consensus among HRV scholars that the role of LF/HF as an index of balance between sympathetic and parasympathetic system is ambiguous, thus lowering its predictive value [11].…”

“…With advances in technology and computer science, HRV analysis has been shifting toward a more predictive approach rather than an explanation-focused strategy using traditional statistical modeling [14]. The development of stress detection systems using HRV indices and machine learning techniques has shown moderate to good accuracies both in laboratory [15,16,17,18,19] and field settings [20,21,22]. This approach has been successful in differentiating non-stress and stress conditions that were commonly induced by various acute stressors such as arithmetic tests, Stroop tests, trier-social threat tests, Montreal imaging, and horror viewing (e.g., [17,18,23,24].…”

“…The development of stress detection systems using HRV indices and machine learning techniques has shown moderate to good accuracies both in laboratory [15,16,17,18,19] and field settings [20,21,22]. This approach has been successful in differentiating non-stress and stress conditions that were commonly induced by various acute stressors such as arithmetic tests, Stroop tests, trier-social threat tests, Montreal imaging, and horror viewing (e.g., [17,18,23,24]. Concepts of stress, emotions, and cognition load overlap and have complex interrelationships [7].…”