Design And Implementation Of Family Smart Healthcare System Based On Smart Terminal

With the problems in healthcare industry become increasingly serious,Chinese thirteenth Five-Year Plan put forward to develop smart healthcare to improve the operation efficiency and service level in healthcare industry.At present,the smart healthcare is still in the early stage of development,the combination between theory and practice in smart healthcare is not enough.At the sametime,the application of smart healthcare is limited to the large medical institution,and it is not applied to our daily life.Therefore,the in-depth integration of smart healthcare and daily life will have great practical value.Aim at the health demand of old groups and health sensitive groups,this thesis design and implement a fimaly smart healthcare system based on Android platform.The main contents are as follows:1.Complete the design and implementation task of the family smart healthcare system and the system implements the functions of health monitoring,fall detection,health prediction and remote consultation on smart terminal,which provide an efficient way for users to manage their health condition.2.Aim at the phenomenon of the olds fall easily,this thesis design a fall detection scheme based on the acceleration transducer to detect fall condition of the olds intelligently.In this scheme,feature extraction and analysis are carried out on the accelerometer data firstly,and then the fall detection model is builded by decision tree C4.5 algorithm,and the test result for this model show that the precision of the fall detection model is up to 91.25%.On the issue of threshold selection in continuous feature,K-means clustering algorithm is applied to optimize the threshold selection,which is useful to select a appropriate threshold.3.Aim at the demand of positioning after fall detection,this thesis designs an optimized indoor positioning schema as the supplement of Baidu positioning,which optimize the precision of indoor positioning in some area.In this schema,the precision of indoor positioning is optimized from the aspects of off-line training and on-line positioning in fingerprint positioning algorithm.Compared with k-nearest neighbor algorithm and weighted k-nearest neighbor algorithm in fingerprint positioning,the precison of the optimized positioning scheme is increased by 20%,and the time complexity of positioning process is reduced from O(n)to O(logn).4.Aim at the demand of health prediction,this thesis designs a modified grey model prediction algorithm to predict the future health data,and make residual test to test the accuracy of the algorithm.The result show that the algorithm is useful in health predictionAs a smart healthcare system which is used in daily life,the system in this thesis implements the function of monitoring health data remotely,detecting fall condition intelligently,predicting health trend intelligently and consulting docotor remotely,which provide an efficient way for users to manage their headth condition.