Estimating Mental Fatigue Based On Physiological Signals Detection And Analysis

There are many theories on fatigue, but its uniform definition has’t been built yet. In some specific area fatigue could be classified as physical fatigue (also called muscle fatigue), mental fatigue (also called psychological fatigue) and mixed fatigue, which is usually called psychophysiological fatigue. Physical fatigue refers to the decline of the muscle’s ability to endure burden, while mental fatigue is defined as the drop of attention, reaction and alertness caused by lack of sleepiness, long working time, high concentration, repeated movement and monotonous environment. Classical mental fatigue includes driver fatigue in transportation (e.g. car, train, plane, ship and so on) after long time driving. The paper mainly pays attention to the mental fatigue.There are many researches on the evaluating of mental fatigue these years, including the subjective way based on the subject’s self-reported fatigue levels and questionnaire, the way according to performance of vehicle such as lane deviation, steering angle and gas pedal, the mechanical vision way by identify physical symptoms such as head nodding, eyelid movement and facial expression with camera, and the way in view of electroencephalogram measure. Because of various disadvantages, they are difficult to be available in practical use. This paper is to find an available and practical method to estimate mental fatigue, which is suitable for the monitoring in mobile situation (for example driving). Given all that, this paper put forward a method, different from the traditionally morphology-based way, which goes deep into subject’s physiological level. However, among so many physiological parameters, the question is which should be chosen for us to study. In this paper we selected parameters of electrodermal activity(EDA), heart rate(HR) and heart rate variability(HRV) for study, based on the principles of noninvasiveness, availableness and practicality.The electrodermal activity(EDA) which includes the tonic activity level and the phasic responses could be recorded by skin resistance or skin conductance. The tonic activity is believed to reflect vigilance and sustained attention, while phasic responses reflect the sudden impact a discrete stimulus may have. In vitro, the conductivity of the skin is wake, but in vivo, it is caused by sympathetic nervous system activation, which changes the levels of sweat in the eccrine sweat glands and controlled by central nervous system. Skin conductivity has been found to be one of the most robust and sensitive non-invasive physiological measures of autonomic nervous system activity and mental situation.The rhythmic activity of heart is to maintain arterial blood supply. In order to keep its steady, the heart activity is affected by complex factors, including the higher nervous activity of brain, the spontaneous rhythm of central nervous system and the reflex activity of cardiovascular system leaded by pressure and chemistry receptor, but in fact, the final expression is the balance between sympathetic system and parasympathetic system. The sympathetic stimulus can strengthen the contraction of myocardium, which accelerates the heart beat, while the parasympathetic stimulus inhibits the activity of the heart to attain the balance. In conclusion heart rate is an overall index to reflect physical and mental stress level in different situations. Heart rate variability refers to the variation in the normal beat-to-beat interval. The normal heart is proved to be a chaos system. The study on the sinus rhythm of heart finds that the R-R intervals of the healthy in rest state show chaotic characters and that of those sick tend to tidy, that is the complexity decreases. Heart rate variability originates from the synergy of the two branches of the autonomic nervous system, so it is an effective indication to reflect the autonomic nervous system. Since the regulation of the autonomic nervous system is affected significantly by the changes of physiological and/or psychological state, one of the key points of this paper is to explore whether the heart rate variability is associated with mental fatigue.The main research work of this paper is as follows:a) Choose electrodermal activity, heart rate and heart rate variability as the physiological parameters to study mental fatigue and design a prototype, which can send real-time data to personal computer for processing, analyzing and displaying, to detect them synchronously;b) Put forward a method to judge mental fatigue by measuring the reaction time and design a PC based human-computer interaction software for testing reaction time;c) Choose 12 healthy young males to do the reaction time test on the PC software, detect their skin conductance and ECG meantime, process the digital signals;d) Statistic results are performed with software SPSS 13.0, expressed in the form of X±S. Using paired t-test to analyze the difference between the sober and mental fatigue state, P<0.05 means the difference is significant.The research results and discoveries are as fellows:a) During the process from sober to mental fatigue state, the overall body function is on the decline, the sympathetic nervous activity becomes lower and the parasympathetic nervous activity becomes higher;b) During the process from sober to mental fatigue state, the detected average skin conductance level, average heart rate and frequency domain parameters of heart rate variability, including total power density, percentage of the very low power density, high power density, percentage of the high power density and ratio of low power density to high power density are all found significantly different (their t values are 2.734,1.653,-2.468,-2.481,2.795 and P values are 0.019,0.033,0.031,0.031,0.017 respectively), while more studies are required to other parameters;c) It’s hopeful to monitor human being’s mental fatigue validly by measuring electrodermal activity, heart rate and heart rate variability jointly, and analyzing trends of their relevant parameters including average skin conductance levels, average heart rates, total power density, very low power density, high power density and ratio of low power density to high power density.