Mobile devices, such as smartphones, have become commonplace in health care settings, leading to the development of both platforms and applications for health care, e.g., HealthKit on iOS, where apps can collect users’ health and activity data and the data will be used for medical research to bring more powerful health solutions. However, no data is collected for the research of infectious diseases. Moreover, currently, most health data are collected by manual input or external devices. Nevertheless, mobile devices equipped with various sensors have the capability to sense data. Therefore, the aim of this research is to provide useful data for the research of infectious diseases, and the focus is to design, implement and deploy mobile systems that can sense related health data automatically.
Human Contact Networks
Infectious diseases pose a serious threat to public health due to its high infectivity and potentially high mortality. Since many infectious diseases spread among people by droplet transmission within a certain range, it is very important to know human contact networks to study infectious diseases. Therefore, we designed a sensing system and deployed it in a high school to collect contacts happened within the disease transmission distance. The system consists of stationary nodes and mobile nodes that can determine whether a contact happened indoor or outdoor, which is important to accurately model disease propagation since the activity of many infectious viruses (e.g., influenza virus) varies in indoor and outdoor environment due to the different ambient airflow patterns.
Based on the collected contacted, a human contact network is constructed to model the disease propagation and we further investigate the problem of targeted vaccination, i.e., vaccinating a small group of people rather than all the people in a community to prevent the spread of a disease. We proposed a metric - connectivity centrality to find the important nodes in the constructed network for disease containment. Connectivity centrality considers both a node’s local and global effect to measure its importance in disease propagation. Centrality based algorithms are presented and further enhanced by exploiting the information of the known infected nodes, which can be detected during targeted vaccination.
SymDetector
Respiratory symptoms are related to illnesses, infections or allergies. Among such symptoms, sound-related respiratory symptoms, such as sneeze, cough, sniffle and throat clearing, are commonly observed and useful in health-related research. However, self-reporting, which has been commonly used in the current research to collect respiratory symptom data, has been shown to be inefficient and inaccurate in recent studies. Therefore, we propose SymDetector, a sensing system built on off-the-shelf smartphones to help researchers collect accurate sound-related respiratory symptom data from users.
SymDetector leverages the built-in microphone sensor to sense the phone’s acoustic context and detect the user’s acoustic events which are related to respiratory symptoms, including sneeze, cough, sniffle and throat clearing. SymDetector can work in an unobtrusive way to collect users’ symptoms for a long period while preserving users’ privacy. SymDetector consists of four components. Audio Sampler reads audio samples from the microphone and segments them as frames and windows. The windows which may potentially contain respiratory symptoms are sifted out by Symptom Detector and fed to Symptom Classifier, where acoustic features are extracted and multilevel classifiers are used to classify the respiratory symptoms. The detection results are then recorded in Symptom Recorder. We have implemented SymDetector on as an Android app and evaluated its performance in real experiments involving 16 users and 204 days.