The collaborative research project developed and evaluated a mobile sensor called EasySense that can provide continuous physiological monitoring without skin contact in the field environments using radio frequency (RF) probes. This approach addresses the problem of physiological monitoring today that requires skin contacts such as electrodes for ECG, and hence cannot scale to widespread monitoring of patients and healthy adults for years. The key challenge is to develop high-resolution sensing on low-power mobile platforms that can separate out the weak motion signals of heart and lung, from the gross motion of the body and the sensor.

The project is developing theory and design for a compressive ultrawideband (UWB) RF sensor that achieves two orders of magnitude reduction in the required sampling rate to make it feasible to realize in a low-power mobile form factor. EasySense employs dynamic compressive sensing algorithms to improve the quality of sensing through temporal integration of information and employs interference subspace cancelation methods to cancel out motion artifacts using data obtained from accelerometers and gyroscopes. The project is implementing all the needed hardware, firmware, embedded software on the sensor node for sampling, processing, and wireless communication, and mobile phone software for data collection, storage, and visualization.

EasySense is evaluated against traditional physiological sensors via lab and field studies on human subjects involving stress and exercise protocols. By realizing contactless sensing of physiology in the field environment, EasySense will enable long-term physiological monitoring at large-scale that is essential for determining potential causes and early biomarkers of fatal diseases of slow accumulation such as cancer and cardiovascular diseases. In addition to being used widely in health research and practice, EasySense can be used for hands-on demonstration in health education. Information on the project, developed hardware and software design files and code relating to the testbed infrastructure will be accessible in open source form via the project web site (http://www.easysense.org).

Grant/Award: NSF-1231754
Lead PI: Dr. Emre Ertin, The Ohio State University