Bruce Schatz, College of Medicine, University of Illinois
Qian Cheng and Joshua Juen, College of Engineering, Institute for Genomic Biology
GaitTrack is a phone app to detect health status, while the person is simply carrying it during normal activities. No additional devices are necessary, only smart phone software. The GaitTrack software monitors gait speed, the walking patterns. It uses the phone sensors, accelerometer and magnetometer, to record spatial temporal motion of the phone. Our software transforms the phone motion into body motion. The phone may be carried anywhere, with activities not normal walking ignored, such as climbing stairs or riding cars. The sensors are only used when the phone is in motion, saving enough energy for continuous monitoring.
Watching how a person walks is the most important part of the neurological examination. A timed walk test is the major assessment for COPD (Chronic Obstructive Pulmonary Disease), a lung disease affecting millions of persons. Along with CHF (Congestive Heart Disease), where walk tests are also major assessments, such diseases are major costs for Medicare and major reasons for rehospitalization. Walking patterns are potential monitors for personality testing, of clinical depression and anxiety, and for joint disease, such as rheumatoid and osteoarthritis.
We have evaluated our software in a pilot study with 30 COPD patients, with IRB approval through the University of Illinois Hospital and Health System. This demonstrates that midrange smartphones can accurately measure gait speed, with 10% error comparable to that in walk tests. We have analyzed medical accelerometers, the current method for free-living walk-assessing, showing that phone sensors with appropriate software are more accurate, since they also sample more frequently than human motion and can measure only during good walking periods.
The GaitTrack software is in research prototype, but this technology can transform every smart phone into a health monitor, to assess status through gait speed. There is immediate application to chronic disease of lung and heart, such as COPD and CHF. A health monitor system using simple phone app can detect variation, individual change in activity status, and assess severity, comparing changed status to normal baseline for a particular level for the chronic disease.
GaitTrack: Health Monitoring of Body Motion from Spatio-Temporal Parameters of Simple Smart Phones
Qian Cheng, Joshua Juen, Yanen Li, Valentin Prieto-Centurion, Jerry A. Krishnan and Bruce Schatz
ACM Conference on Bioinformatics, Computational Biology and Biomedical Informatics, Health Informatics Symposium, September 2013