Diagnostic Quality Driven Energy-efficient and Bandwidth-scalable Physiological Data Collection for Wireless Telehealth

Smartphones are fast emerging as the platform for wireless healthcare via affordable medical sensors. However, the requirement of high rate physiological data collection has precluded successful large-scale deployments due to shortened battery lifetime and lack of bandwidth scalability issues. It is thus essential to develop mechanisms for reducing the amount of data that needs to be transferred while maintaining diagnostic accuracy by preserving specific signal features. We believe that while collecting data from a sensor, the system should not uniformly adjust the degree of precision or compress the data. Instead, it should reduce the amount of data in a fashion that preserves "important" signal features that are enough to classify a context (make the diagnosis) with a required confidence level. We demonstrate that the proposed approach provides accurate inference results while effectively reducing the communication load.