Design Of Data Processing Platform And Its Algorithms For Wearable Multiple Physiological Parameters Monitoring

Although wearable devices has played a part role in safeguarding of human health,but their shortcomings restrict their progress and growth in health care domain.Therefore,a monitoring system of wearable multiple physiological parameters is implemented in this thesis.The system contains five parts,a wearable physiological parameter acquisition module,an APP on Android terminal,back-end server,webpages and physiological parameters processing algorithms.The duty of the acquisition module is not only data collection of electrocardiogram(ECG),acceleration and shell temperature signal but also data transmition using Bluetooth low energy technology.The APP on Android terminal can receive,display,save and upload the data.A running program on backend server process the data of physiological parameters and response to the web page request.Users' physiological parameters can be managed in the webpages.In this thesis,the physiological parameter processing algorithms are composed of arrhythmia classification and human activity pattern recognition.In arrhythmia classification,First,ECG signals are the time-frequency features are extracted by Stransform.Second,the genetic algorithm and support vector machine(SVM)are combined as a Wrapper approach to search an optimal feature subset.The feature weights are computed by ReliefF,and the initialization of genetic population depends on the feature weights.The genetic algorithm searches an optimal feature subset using classification performance as the fitness function.Finally,a Multi-SVM model using one against all strategy is built for the classification of 8 types of ECG beats from the MIT-BIH arrhythmia database.Experimental results indicate that the proposed approach has the best performance among other state-of-the-art approaches.In human activity pattern recognition,time and frequency domain features are extracted,and the feature selection method and classification model in arrhythmia classification is fully utilized.Then 10 types of human activity patterns are recognized.In addition,the function of monitoring physiological parameters is implemented with two algorithms program running in back-end server.