People around the world increasingly use their mobile devices to track health indicators and report the results to their doctors. Zion China, a provider of mobile healthcare and telemedicine services, is a part of this trend. The Beijing company has developed innovative health monitoring and analysis tools that collect data on blood sugar, diet, exercise, and medication for each patient and resulting personalized advice to help patients manage their health. Its developers realized that as the company’s installation base grew, the sheer volume of real-time data presented a scalability and data management problem. They collaborated with Microsoft to take advantage of the scalability of the Azure IoT Suite, Machine Learning, Stream Analytics, SQL Database, and Microsoft Power BI while also improving analytics and data security.
At the heart of why Zion China built this solution is that "China currently has 130 million citizens who have been diagnosed with diabetes," says Vincent Yang, CEO of Zion China. "It is our mission to help these patients and their families manage this chronic disease."
Zion China’s original technical solution, E-Followup, was based mostly on traditional BI with data sourced from on-premises and various devices or cloud storage. In this engagement, they wanted to achieve these technical and business goals:
1. Design and architect a smart, fast, and cost-effective way to continuously feed data from devices to the cloud
While wearing the device, a user will generate more than 3,000 readings of blood glucose in a 7-day period, once every 3 minutes. At the same time, the device collects data on the user’s daily diet, exercise, medication, insulin, and other information. High-frequency data transmission and large amounts of data storage have become a technical problem. The newly designed architecture addressed the ingest of data from the devices into Azure to build a robust solution.
2. Optimize the analytic suite, moving from BI analysis to proactive prediction to get insights from data easier
In the past, even with a data analytic product, doctors spent 4 hours or more on each data analysis to identify the impact of various factors on the user and find the correlation between the data. Users and doctors need to view data in many different scenarios, such as web pages, mobile apps, tablets, and so on. Zion China needs a tool that can be adapted to multiple devices and is easy to develop and proficient in data analysis and presentation.
3. Persist data storage for future use and ensure data security
Zion China wants to ensure that all the persisted stored data, even with all personal information removed, is stored in a highly secure manner.
Solution to handling massive data volumes
Xi Wang, Microsoft Technical Evangelist, says, "For any given patient, E-Followup transmits huge volumes of data over time. Making that transmission more efficient, and improving the ability to scale the storage and analysis of the data, required a new architecture," explains.
The Microsoft China team worked with Zion China to re-architect the solution:
- Adding IoT Hub to the front end to serve real-time data transmission from device to cloud.
- Using Azure Machine Learning to generate proactive prediction on glucose and patient data.
- Using Transparent Data Encryption in Azure SQL Database to secure data.
- Using Power BI and Power BI Embedded for easier visualization of insights.
New Architecture
Key technical components included in the solution are:
Next steps
Zion China is expanding its solutions to top hospitals and healthcare management organizations. With the new IoT Hub, Machine Learning, and the full stack powered by Azure, Zion China can expand its business and be more agile and cost-effective.
Get the code and architecture from this project on GitHub, watch the interview with the Zion China and Microsoft team, get hands on with IoT labs or start to build your own IoT solution on Azure.
