Measuring the Complexity of EMG Signal by Using Fuzzy Approximate Entropy in Post-Stroke Patients Rehabilitation

Stroke is one of world’s highest medical cases that cause physical or mental problem, even death. The treatment for overcoming stroke so far was by conducting rehabilitation or therapy. The therapy progress could be seen over the improvement indicated by stroke patients after participation in rehabilitation program. So far, the assessment of the level of development exhibited by stroke patients are still subjective. Meanwhile, Electromyograph (EMG) can be used as a quantitative method to evaluate the patient's improvement. The complexity of the EMG signal could define the progress of a rehabilitation process related to muscle function. One of the methods to calculate the complexity is fuzzy approximate entropy (fApEn). This study aims to propose fApEn as a quantitative method to evaluate the condition of stroke patients by comparing the complexity of the normal EMG, affected, and unaffected side limb of the patient's EMG signal. The value of fApEn on Subject 1 showed that the affected side obtained the complexity of 0.18141 and on the unaffected side of 0.80740. Subject 2 showed the complexity on the affected side of 0.58597 and the unaffected side of 1.02787, whereas the healthy subjects showed the complexity of 2.07856. The statistical results showed that the data were homogeneous and normally distributed. There was a significant difference between the results of fApEn on the affected side and the unaffected side. Thus, fApEn could be used as a quantitative method to determine the condition of muscle activity.