The Design Of A Real-Time System For Continuous Monitoring Of Parkinson’s Tremor

Currently, Parkinson’s disease is the second most common neurodegenerative disease, the average age of patients initially diagnosed with this disease is62. Larger number of people will suffer from this disease along with the expected growing of elder population over the next20years. Unfortunately, so far, there is no effective treatment for this disease. However, an early diagnosis of Parkinson’s disease may result in the delay of severe motor disturbances by some health management policies.Early diagnosis of Parkinson disease requires proper quantization of tremor, which usually occurs on patient’s fingers or hands. This tremor is divided into three types:resting tremor, posture tremor, and action tremor, among them, resting tremor is the most common one. Currently, clinical tremor diagnosis mainly relies on subjective methods, such as UPDRS scales. Therefore, lots of researchers have devoted their efforts to the accurate tremor quantization by using time or frequency domain analysis of the signal collected by inertial sensors and other equipment. More specifically, accelerometers have been widely used for the assessment of Parkinson’s tremor or essential tremor. Laser lines with image sensing and electromyography are also used. Different signal analysis methods have proposed by previous researchers such as power spectrum analysis and wavelet analysis in their researches. The objective of this study was to develop a3-D accelerometer based real-time tremor severity monitoring system for Parkinson’s patients. A compound tremor severity index calculated by considering average power in time domain and maximum peak in frequency domain of the tremor signal was introduced. The distribution of different tremor severities for certain period of time can be plotted graphically. The proposed system with tremor severity identification algorithm has been verified by Parkinson patient’s tremor signals. In addition, the robustness of the system was evaluated with healthy participators performing6activities with imitated tremor symptom. Experimental results show that the proposed system may become an effective tool in tremor quantization and severity identification for Parkinson’s disease patients.