A surface electromyography-based pre-impact fall detection method

Falls and fall related injuries seriously threaten the health of the elderly. To deal with the problem, a human pre-impact fall detection method based on surface electromyography (sEMG) signals was proposed. 20 subjects were recruited to collect the lower limb sEMG signals of their activities of daily living (ADLs) and fall process. The motion capture system was used to collect the coordinates of marker points in the subject's motion in real time. Time-domain features of the 4-channel EMG signals were extracted to construct feature vectors. Then Support Vector Machine(SVM) was trained and the activity was identified using the obtained classifiers. The experimental results showed that the sensitivity of the method reached 93.71%, the specificity reached 92.67%, and the average lead time was 202.4ms. This method can effectively predict falls and distinguish them from ADLs.