Excellence in Research: Privacy preserved data fusion for personalized recommendation in an Edge-assisted IoT smart health network

Remote patient health monitoring systems have been around for quite a while but only recently did the concept of smart health monitoring systems gain popularity. With the deterioration in people's health and increase in mortality rate, researchers are always looking for ways to provide more support to the patients by improving critical delays in responding to a life-threatening event, incorrect medications, and day-to-day improvements that can be done to elongate patient's health. Given that the numbers of devices connected to the internet by the people are growing rapidly, health insurance companies see this as an opportunity to help people with their care. The rise of wearable and implantable devices in healthcare provides patients and providers with remote diagnostic tools, further reducing the need for expensive care. This project will focus on combining data from various sources, often referred to as data fusion, to provide patients with more accurate prescriptions, treatment, and suggestions to improve their health.The goal of this project is to investigate and develop a framework that collects and merges data from various smart health sensors to predict user behavior and alert concerned authorities for emergencies. This framework allows the server to collect the context of the user health without violating user privacy. Researchers would use not only user's health sensors that are used on a daily basis but also appropriate data from hospitals, labs, and pharmacies to benefit the smart prediction system. The specific objectives are to do the following:1. Design and develop an application to collect sensor data and merge external databases2. Design and develop framework to address missing values in the data3. Design and develop framework to input context information in a privacy enhanced fusion4. Investigate the trade-off between context-aware fusion Vs non-context-aware fusion5. Design and develop framework to address data dependencies for user behavior and generate personalized results.The project also heavily depends on the involvement of undergraduate and graduate students. Project-based learning approach will be followed to integrate research and education for the students. Courses taught in the department, including machine learning and big data, and development of new courses, including mobile application development and Introduction to IoT, will serve as significant contributors towards students learning.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.