Achieving Smart Health with Mobile Devices

Restricted (Penn State Only)
Sun, Xiao
Graduate Program:
Computer Science and Engineering
Doctor of Philosophy
Document Type:
Date of Defense:
September 28, 2016
Committee Members:
  • Guohong Cao, Dissertation Advisor
  • Guohong Cao, Committee Chair
  • George Kesidis, Committee Member
  • Sencun Zhu, Committee Member
  • Siyang Zheng, Outside Member
  • Smart health
  • Mobile sensing
  • Mobile computing
  • Machine learning
  • Disease containment
Existing healthcare systems are cost-prohibitive and inefficient, and thus there is an urgent need for the transformation of healthcare from expensive, reactive and hospital-centered to low-cost, proactive and person-centered. The rapid development of mobile devices over the past few years has opened up opportunities for smart health, bringing an inexpensive, convenient and efficient way for medical data collection and health monitoring. With various built-in sensors, advanced processors and wireless communication capabilities, mobile devices can be exploited to develop a broad range of healthcare applications. The specific goal of this dissertation is to design systems for achieving smart health with various mobile devices (e.g., smartwatch, smartphone and wireless mote). First, we design a smartwatch based system which leverages the built-in accelerometer to monitor the respiratory rate and body position during sleep. To calculate respiratory rate, we design a filter to extract the weak respiratory signal from the noisy accelerometer data collected on the wrist, and use frequency analysis to estimate respiratory rate from the data along each axis. Further, we design a multi-axis fusion approach which can adaptively adjust the estimates from the three axes and then significantly improve the estimation accuracy. To detect the body position, we apply machine learning techniques based on the features extracted from the accelerometer data. Second, we present another smartwatch based system which detects 8 common daily activities, including sitting, walking, running, going upstairs, going downstairs, eating, driving and sitting in a vehicle. By leveraging the built-in motion sensors on the smartwatch (i.e., accelerometer and gyroscope), a multi-level classification system is proposed which considers both detection accuracy and energy efficiency. Several practical issues are considered in developing the system to meet the specific requirements of the smartwatch, such as sensing capabilities and resource constraints of the smartwatch, and different wearing habits of smartwatch users. Third, we propose a smartphone based system to unobtrusively detect the sound-related respiratory symptoms occurred in a user's daily life, including sneeze, cough, sniffle and throat clearing. The built-in microphone on the smartphone is used to continuously monitor a user's acoustic data and uses multi-level processes to detect and classify the respiratory symptoms. Several practical issues are considered in developing the system, such as users' privacy concerns about their acoustic data, resource constraints of the smartphone and different contexts of the smartphone. Finally, we design targeted vaccination for infectious disease containment. To understand the disease propagation better, we collect student contact traces in a high school based on wireless motes carried by students. With our wireless system, we can record student contacts within the disease propagation distance, and then construct a disease propagation graph to model the infectious disease propagation. Based on this graph, we propose a metric called connectivity centrality to measure a node's importance during disease propagation and design centrality based algorithms for targeted vaccination. Further, we enhance our solution by exploiting the knowledge of the infected nodes which have been detected during vaccination.