Muscle activity obtained using electromyography (EMG) sensors has been used in previous studies for the evaluation of load on the lower back muscles. However, there is less research on the relationship between human perception of changes in lower back muscles and EMG signals. In this study, the just noticeable difference (JND) method is used to determine the minimum change in muscle exertion that humans can perceive, as quantified through four EMG amplitude-based metrics: mean, median, 90th percentile, and root mean squared (RMS) values. Fourteen participants performed stoop bending tasks with varying loads (0.5 kg to 4.5 kg) on the lower back. After each pair of trials, a pairwise comparison was conducted to ask participants which load resulted in higher exertion on the lower back. EMG data was collected from lower back muscles to calculate the four metrics. Using these metrics and participants’ feedback, JND values for each metric were calculated. The results showed the average JND values of 29%, 37%, 27.0%, and 40% for mean, median, 90th percentile, and RMS of EMG amplitude, respectively. Analysis shows no statistically significant differences among JND values of different metrics, except for the 90th percentile. The 90th percentile showed statistically significant differences compared to the median and RMS; however, statistical power was limited (0.52–0.58), and the findings should be interpreted as preliminary and require replication with larger samples. Our findings highlight the importance of designing exoskeletons that provide perceptible reductions in muscle activity, while considering perceptual differences, to enhance user acceptance.
Zindashti et al. (Mon,) studied this question.