A Wearable Mobihealth Care System Supporting Real-time Diagnosis And Alarm

Obejectives: Wearable technology is a non-invasive effective algorithm to record long-term physiological signals. Wearable telemedical instrument will be important device for monitoring, diagnosing, and health care under new medical system, its broad application will promote the development of telemedical system, home healthcare system, and individual healthcare system. This research is aiming at realizing long-term monitoring and real-time diagnosis for physiological parameters (ECG, respiration, activity, and body temperature), and providing remote transmission of vital signs, GPS, and alarm in case of emergency situation.Methods: Firstly, a new real-time ECG diagnosing algorithm is presented based on other fruits: ECG beat classification algorithm based on combination of morphological template matching and characteristic extraction. The algorithm improves old morphological template matching algorithm, builds multiple morphological templates, uses standardized matching threshold and noise evaluation to assist ECG beat template matching analysis. Based on the result of beat matching, characteristic extraction algorithm is used to provide rapid ECG beats classification. Secondly, at aspect of real-time human fall detection, the synthesized determinant algorithm of linear fall and nonlinear fall based on triaxial accelerometer is presented. Synthesized threshold determinant algorithm of multiple uniaxial accelerometers and threshold determinant algorithm of total accelerometer can detect linear fall and nonlinear fall respectively, detecting thresholds can be acquired by experiment, so real-time human fall detection can be realized. Thirdly, by absorbing designing notion of wearable technology and integrating other physiological signals detection technology into elastic T-shirt, WIMS with short range wireless communication is developed. At the same time, developing PPU with the function of short range wireless communication, telecommunication, wireless GPS, and real-time diagnosis, together with MSC and WIMS, constitutes wearable mobihealth monitoring system. Finally, the feasibility and reliability of the system will be test by medical experiment.Conternts: The following works have been performed in this paper:①Noble bio-potential pre-amplify circuit is studied. Textile electrodes with bio-potential character are used substituting for traditional gel electrodes. High inherent impedance and skin-electrode impedance of textile electrodes bring higher requirements for ECG amplification circuit. Active electrodes, bootstrap amplifier, driven-right-leg circuit, and adding hydrogel membrane are used to resolve this problem.②The ECG preprocessing and basic characteristic points detection are studied. The ECG preprocessing includes the restraint of high frequency noise, baseline wander, and power line interference. The basic characteristic points detection includes P wave, T wave, QRS wave, onset, and offset point detection.③Real-time ECG beat classification algorithm is studied in detail. Considering the problems existing in current algorithms, a new real-time algorithm fitting microprocessor: ECG beat classification algorithm based on combination of morphological template matching and characteristic extraction, is presented, and it has been implemented in ARM7. With the evaluation test with MIT-BIH proved, its specificity and sensitivity for PVC reach 93.38% and 96.69% respectively. Not only accuracy, but also the speed of the algorithm is satisfying.④Human activity detection technology, especially human fall detection technology is studied. Considering the disadvantage and shortage of current fall detection algorithms, the fall detection algorithm with the combination of linear fall and nonlinear fall is presented.⑤"A wearable mobihealth care system supporting real-time diagnosis and alarm"is developed based on the new algorithms studied by integrating wearable technology, physiological signals detection technology, wireless communication and GPS technology. This paper describes the software and hardware development process of the three part of the system: WIMS, PPU, and MSC, especially the realization of real-time diagnosis algorithms implemented in microprocessor, real-time alarm mechanism, and feedback mechanism of health status.⑥A medical experiment was designed to check the feasibility and reliability of the system.Achievements: New real-time ECG diagnosing algorithm and fall detection algorithm are presented, and they are implemented with C language in ARM7 processor;"Wearable mobihealth care system supporting real-time diagnosis and alarm"was built based on CDMA by integrating wearable technology, physiological signal detection and processing technology, gpsOne technology, and wireless communication technology; A prototype was developed, the feasibility and the reliability of the system were validated by medical experiment for 15 volunteers, and the system was positively rated; According to the design of wearable mobihealth monitoring system and real-time fall detection system, a patent of State had been applied for respectively.Conclusions: This system absolutely exceeds conventional medical system, realizes non-intrusive long-term monitoring, real-time diagnosis, and abnormal alarm, and realizes dynamic monitoring, mobile monitoring, and tracing monitoring; Real-time online diagnosis for physiological signals are realized. On the one hand, timely and accurate abnormal alarm can safeguard the patients. On the other hand, building real-time data traffic connection with monitoring center can be avoided, which will greatly reduce data stream and cut down cost;"Wearable mobihealth care system supporting real-time diagnosis and alarm"has good performance, worn comfort, and easy to use, which can be used to record long-time vital signs in low physiological and psychological variables. It can also be widely used in military medicine and biomedical engineering study.The innovations of this study are as followings:①High integrated wearable physiological signals detection/monitoring system realizes non-invasive detection of multi-parameter physiological signals in dynamic condition.②A novel real-time ECG beat classification algorithm is proposed: ECG beat classification algorithm based on combination of morphological template matching and characteristic extraction.③A novel real-time human fall detection algorithm: the fall detection algorithm with the combination of linear fall and nonlinear fall based on triaxial accelerometer sensor.④Real-time diagnosis and alarm of Multiple physiological parameters: Real-time ECG beat classification algorithm and human fall detection algorithm were realized in ARM7 microprocessor.⑤Separation structure between PPU and WIMS: PPU is separated from WIMS, wireless communication is applied with. This makes PPU more easy to use and look, and avoids placing boring leads.⑥Integrating advanced communication and GPS technology: The gpsOne and CMDA technology integrated into system were able to provide whole region tracking monitoring for patients.⑦Signal processing and data mining technology with dynamic data: Long-term physiological data from patients in daily living condition can be acquired through our study, which will be impossible for conventional monitoring devices, thereby more in-depth study can be performed with signal processing method.