Research And Development Of The C/S-based Remote Service System For Noninvasive Blood Glucose Detection

Mobile Health is an emerging field which combines mobile Internet and medical services. Traditional health care model needs to be reformed, and mobile health acquires widespread concern of academia and industry, by making medical diagnosis based on data accumulation and integrating medical service supply chain. However, medical equipment requires high accuracy physiological sensor dues to the characteristic of portability in practice. Useful informat ion needs to be extracted from low SNR signals during data processing, and equipment consistency differences should be effectively shielded. Meanwhile, the model and algorithm used for recognition of human physiological characteristics and adapted for motion detection device are not perfect. There is a big accuracy difference between professional medical equipment and mobile testing equipment. The individual differences of physical characteristics are also required to provide personalized mobile medical healthcare solutions.Mobile medical is applied to non-invasive blood glucose monitoring and diabetes prevention, through building a non-invasive blood glucose monitoring remote service system which characterized by "Client-Cloud-Data". Firstly, the STM32 implementation for the non-invasive blood glucose monitoring module was rewritten, and the error source of the module’s hardware was analyzed. An improvement program was proposed, and product consistency improvements were verified. The software system of integrated and split detector were designed and developed based on the Android system. At the level of "client", the sensor error was reduced, the consistency of the product was improved and the user experience of non-invasive blood glucose monitoring was enhanced. Secondly, a LAMP architecture server based on Ali cloud IaaS platform was built. Medical service application used Struts + Spring + Hibernate architecture implementation. Based on this implementation, the interface was designed to connect end-user, medical institutions and pharmaceutical suppliers. Web Service was built for managing system services. A unified external service can be provided through shielding platform differences. Finally, based on Bergman minimization model, compensation program of impact of the interstit ial tissue insulin acts on non-invasive blood glucose monitoring was proposed, through the hospital experiment and online collection, the non-invasive blood glucose monitoring accuracy and mobile medical personalization based on the classification algorithm were improved at the level of "data".