Quantification Method Of Movement Disorders Based On C-band

With the aging society in China,Parkinson’s disease and other common nervous system diseases of the elderly are increasing,which not only increases the cost of care for medical staff,but also seriously reduces the quality of life of patients.Dyskinesia is the main symptom of Parkinson’s disease.Accurate evaluation of Parkinson’s disease is very important for the correct diagnosis and monitoring of Parkinson’s disease.The gold standard for assessing its severity and that of other movement disorder’s symptoms is the evaluation by a well-trained clinician using standard clinical rating scales.To sum up,clinical evaluation needs to introduce objective scoring method to eliminate the inconsistencies between evaluators,so as to improve the monitoring of motor disorders,and help doctors to develop reasonable treatment plans for patients The purpose of this thesis is to obtain the motion signals of patients based on wireless sensing technology,and to study and analyze the methods of recognition and objective quantification of the symptoms of motor disorders.The main work of this thesis is as follows:(1)This thesis summarizes the current research status of the evaluation of motion disorders and expounds the principle and mechanism of non-contact wireless sensing technology.(2)Establish a wireless sensing platform.According to the actions of the MDS-UPDRS and the lower limb paresis test,a commercial network card was used to obtain the wireless signal disturbed by the subject’s movement.We extract the amplitude and phase information of the wireless signal.Obtain the phase difference between adjacent antennas through the channel quotient to calibrate the original phase data.Outlier filtering algorithm and wavelet threshold filtering algorithm respectively remove outliers and high frequency noise of motion signals.To facilitate feature extraction,a locally weighted regression algorithm is used to smooth the waveform.(3)A one-dimensional convolutional neural network structure based on a convolutional neural network is proposed in this thesis.The neural network model is applied to multiple clinical tasks.It can not only realize the characteristics of dyskinesia signals,but also evaluate the patient’s condition.(4)In the thesis,finger tap and arm rotation were used to quantify the patient’s upper limb dyskinesia,and the lower limb function was evaluated by the lower limb paresis test.The effectiveness of wireless signal quantification of dyskinesia was verified through a large number of experiments.We explored the accuracy of amplitude,phase difference,and amplitude fusion phase difference for different clinical tasks.The results show that the amplitude fusion phase difference is suitable for clinical experiments with a large range of motion,and the phase difference is suitable for clinical tasks with a small range of motion.According to different experiments,the optimal number of wavelet decomposition layers is determined to ensure high-precision classification of patient severity.In summary,the non-contact dyskinesia monitoring system proposed in this paper can effectively obtain the patient’s natural motion signals and can objectively and accurately assess the patient’s condition.To a certain extent,it can replace the current method and improve the monitoring of movement disorders.