Development of Composite Scale of Body Fatigue by Using Electromyography and Subjective Measure during Squat Lifting

Abstract

Work related musculoskeletal disorders affect numerous workers each year. The repetitive tasks sometimes combined with forceful exertions expose workers to risk of over exertion which may deem susceptible of developing musculoskeletal disorders. Among the major tasks being performed in the manufacturing sector, lifting is the most common type of task encountered daily. Researches have adopted different methodologies for assessing the fatigue levels during lifting tasks. The current methods of assessing the occupational exposure to repetitiveness focuses more on the local muscle fatigue and single indicators of fatigue such as electromyography, however fail to provide a meaningful picture of how the multi-dimensional phenomenon of fatigue can be represented as a unified scale representing whole body fatigue. Furthermore, most of the studies focus on a particular region of perceived exertion rather than incorporating the maximum indicators of fatigue to make a comprehensive framework. Therefore an alternate approach of assessment is required that not only incorporates multiple features of the phenomenon of the fatigue but also provides an easy assessment of fatigue through a single scale. This research is an attempt to address the assessment of multidimensional nature of fatigue and introduces a method of incorporating the physiological and psychological variables of fatigue using electromyography, heart rate and Borg scale. The variables are used to assess the level of fatigue and to find how each of the underlying variable contribute towards perceived whole body fatigue. For this purpose, the regression modelling approach is used. The study is carried out in two phases: the design of experiment phase (1) and the scale development phase (2). Each of the phase is carried out to compliment the preceding steps and provide with the relevant data related to the variables selected. The perceived fatigued from the upper and lower extremity of the body and the whole body is obtained using Borg scale while the time series recordings of the muscle activity is obtained with the surface electromyography (sEMG). The electromyography signals of each contraction are identified through root-mean-square (RMS) amplitude which are then transformed from time domain to frequency domain. The RMS is used to find the onset of the contraction and to accurately calculate the mean power frequency at each contraction. The validated Borg scale ratings from the upper and lower extremities are observed and it is found that the lower extremity dominates the perceived exertion during squat lifting. The key muscles undergoing fatigue are found through principal component analysis (PCA). All of the eight muscles of the upper and lower extremities are correlated with their respective extremities to find if there is any relation between the mean power frequencies and the perceived exertions. Heart rate and the whole body perceived exertion indexes are found to have very strong correlation. Logistic regression modeling is used to quantify the measured indicators of fatigue into fatigue and non-fatigue states and with dichotomous responses. The key muscle from the Principal component analysis, the heart rate and the lower extremity are the independent variable, while the whole body fatigue is quantified into dichotomous as well as four levels as a dependent variable separately. For the adequacy assessment, the Borg scale, EMG and Heart rates are selected as the variables being categorized as physiological and psychological indexes. The factor analysis is performed through PCA to have optimum number of the variables for the logistic regression models. The data for each of the variable is checked for reliability, using Cronbach alpha test. In the end the accuracy check for the model is performed. The average accuracy is found to be 80 percent. The model is also checked without incorporating the perceived rating of lower and upper extremities but heart rate is included as independent variable. There was significant difference found between the two methods at higher fatigue level. The dichotomous scale developed through logistic regression modelling will help ergonomists interested in understanding the importance of multidimensional nature of whole body fatigue and modeling the fatigue for other lifting tasks. The methodology presented in this study introduces a new metric for determination of whole body fatigue by incorporating combined factors of fatigue The findings revealed that it is significant to take into account the relation between perceived and measured effort that can help prevent musculoskeletal disorders in repetitive occupational tasks.