Quantitative indices for monitoring mental fatigue based on heart rate variability analysis methods
Sergey E. Burmistrov1, Elena V. Navrotskaya1, Aleksandr V. Kurbako1, Borys P. Bezruchko; 1Saratov State University, Saratov, Russia;
Abstract
Mental fatigue is a significant risk factor leading to reduced productivity and increased operator error probability in critical industries. Existing assessment methods based on subjective questionnaires are unreliable and unsuitable for continuous monitoring. Therefore, the development of objective diagnostic tools using physiological markers is an urgent scientific task. Heart rate variability (HRV) analysis is a promising direction. This study aims to identify the most informative quantitative HRV indices (statistical, spectral, and nonlinear) for diagnosing mental fatigue by comparing their values before and after prolonged cognitive load.
The study involved 25 healthy volunteers (mean age 23.4±2.1 years). The experiment included a baseline psychological screening, a 15-minute resting ECG recording, an intensive cognitive task (serial arithmetic subtraction) until subjective exhaustion (60-240 minutes), and a final post-task ECG recording and psychological testing. ECG was recorded at 250 Hz. A complex of HRV indices was calculated: statistical (HR, D, SI), spectral (TP, LF/HF), and nonlinear (Lyapunov exponent - Lyap, duration of stationary segment in LF and HF bands - DSS-LF, DSS-HF). Statistical significance was assessed using the Mann-Whitney U-test.
Psychometric testing confirmed the development of mental fatigue in all participants. Prolonged cognitive load caused statistically significant (p<0.05) changes in autonomic regulation: a decrease in HR (20%) and SI (47%), and an increase in overall variability (TP increased by 63%, D by 96%). The Lyapunov exponent (Lyap) decreased by 13%, indicating reduced system chaos. However, the LF/HF ratio, DSS-LF, and DSS-HF showed no significant changes.
Thus, the most sensitive biomarkers of mental fatigue were parameters reflecting overall variability (TP, D) and regulatory system tension (SI, HR). Spectral and nonlinear indices characterizing the balance and stationarity of autonomic control (LF/HF, DSS-LF, DSS-HF) did not demonstrate significant dynamics.
Speaker
Sergey Evgenyevich Burmistrov
Saratov State University
Russia
Discussion
Ask question
