Effectiveness Evaluation Based on Electroencephalography and Electrocardiography Analysis for Robotic Therapeutic Massage Application

Therapeutic massage is an ancient technique, which promotes human relaxation and enhances health of human body. It is considered as a part of complementary and alternative medicine. Due to increased demand for healthcare, including medical and elderly cares, many kinds of research dedicate to the development of autonomous healthcare and elderly care systems in resolving the shortage of caregivers and increase the medical quality. Therefore, automated massage system is desirable. Although many researches focus on developing automated massage system, they didn’t measure the biomedical signal of the subjects to prove the effectiveness of their system. This thesis is the first research using an anthropomorphic dual arm robot to develop massage application. Besides, we measure the biomedical signal of the subjects who received our robotic massage as the clues of the effectiveness. An anthropomorphic dual arm robot is developed in our NTU-iCeiRA Lab for therapeutic massage application. In order to evaluate the effects of robotic massage, Electroencephalography (EEG) and Electrocardiography (ECG) signal of the subject are analyzed. Compare to other massage systems, our dual arm robot can provide diversified massage techniques through the impedance control of both Cartesian space and joint space. To evaluate the feelings of relaxation for those subjects being massaged with a more objective indicator, EEG and ECG signal of subjects are measured before and after massage manipulation. In EEG analysis approach, Independent Components Analysis (ICA) is used to filter out artifacts of EEG signals. After signal processing, the power spectrum of delta, alpha and beta rhythms are analyzed. On the other hand, the features of heart rate variability (HRV) of ECG signal, including mean HR, SDNN, RMSSD and pNN50, are extracted and analyzed to evaluate the effectiveness of robotic massage. In experimental results, the variations of EEG signal in power of three different frequency bands are compared with literature which the massage action was conducted by therapist. In addition, the ECG signals are collected both experimental group and control group and the HRV of ECG signals are analyzed. The experimental results show that most subjects felt relaxed after the robotic massage. The results also provide the instrumental evidence of the robotic therapeutic massage effects presented in this thesis.