Analysis of electrophysiological measurements during stress monitoring

Work-related musculoskeletal disorders are a growing problem in todays society. These musculoskeletal disorders are caused by, amongst others, repetitive movements and mental stress. Stress is defined as the mismatch between a perceived demand and the perceived capacities to meet this demand. Although stress has a subjective origin, several physiological manifestations (e.g. cardiovascular and muscular) occur during periods of perceived stress. New insight and algorithms to extract information, related to stress are beneficial. Therefore, two series of stress experiments are executed in a laboratory environment, where subjects underwent different tasks inducing physical strain, mental stress and a combination of both. In this manuscript, new and modified algorithms for electromyography signals are presented that improve the individual analysis of electromyography signals. A first algorithm removes the interference of the electrical activity of the heart on singlechannel electromyography measurements. This interference signal is the major source of contamination in the electrical activity of the muscles in the shoulder and needs to be removed. The algorithm is based on a single-channel extension of independent component analysis to identify statistically independent sources in the signal. The extension consists of the decomposition of a single recording, using ensembled empirical mode decomposition or wavelets and on which traditional independent component analysis is applied. A second data-driven algorithm estimates the rest-activation period from a muscle. Several applications need an accurate estimation of the period when a muscle is in rest. A new approach is presented, using the frequency content of the signals, which is able to distinguish between the rest and the active state of a muscle without a reference measurement or without the unreliable amplitude domain thresholding. The signals of the muscles in the shoulder area showed an interesting specific pattern during the stress experiments after removing the different artefacts. In approximately 65% of the subjects, a continuous firing of a single motor unit, was visible. This indicates a very low and subconscious muscle contraction without a postural benefit. An algorithm to detect these single motor unit firings is presented. The algorithm, based on an energy operator and correlation calculation, showed an excellent performance as it reaches a sensitivity of 100% and a specificity of 94,8%. Another focus of the thesis is the physiological data interpretation during the stress experiments. The muscle activity analysis of the muscles in the shoulder girdle did not reveal any statistically significant increase in muscle activity. An activation of the muscles in the shoulder girdle during the stress test was identified, but this increase was not restricted to the periods where a mental task was executed. Even more, several subjects showed also reactions on other muscles (in the face, the lower back revealing that the muscle activity analysis should not be limited to the trapezius muscle, but broadened to a group of muscles where reaction due to a mental stressor could be expected. The analysis of the heart rate variability, a noninvasive estimation of the autonomic nervous system modulation on the heart rate, showed a statistically significant difference between the different tasks and revealed an extra reduced vagal modulation when the physical and the mental task are combined compared to both tasks seperately. Time-frequency analysis revealed that the effect of the mental task on the physiological signals reduces over time. In an additional study on pregnant women, we were not able to correlate the amount of reaction in the heart rate variability during different conditions with the subjective scores of anxiety and stress susceptibility of the pregnant women. This reveals that there is a clear effect of the tasks on the heart rate variability itself, but this is not one-to-one related with the stress level of the test subjects. Both, the analysis of the muscle activity and the heart rate variability reveal the need for individual analysis during stress monitoring. The stress system consists of very complex interactions in the brain, which are at this moment not completely understood. Those interactions, however, are responsible for the variability in reaction of individual people on stressors: both in the intensity of the reaction as in the physiological reaction itself. A last focus of this thesis was the analysis of the combination of the oxygenation and the electrical activity in the muscle, respectively measured and quantified by near-infrared spectroscopy and surface electromyography, during a muscle fatigue test. The combination of these two signals provides complementary information regarding muscle fatigue. A typical four-phase response was identified in the muscle oxygenation index, where different parameters could be linked with the physiology. A medical application, of which a pilot study is described, where both measurements are used, is the analysis of the progress of the disease in children with a muscle disease (Duchenne Muscle Dystrophy).