E. Sazonov in S.C. Mukhopadhyay et al. (Eds.): Pervasive & Mob. Sens. & Comput. for Healthcare, SSMI 2, pp. 89–110. Springer-Verlag Berlin Heidelberg 2013
Monitoring of posture allocations and activities is important for such applications as physical activity management, energy expenditure estimation, stroke rehabilitation and others. At present, accurate devices rely on multiple sensors distributed on the body and thus may be too obtrusive for everyday use. This chapter presents an overview of a novel wearable footwear sensor (SmartShoe), which is capable of very accurate recognition of most common postures and activities while being minimally intrusive to the subject. SmartShoe relies on capturing information from patterns of heel acceleration and plantar pressure to differentiate weight-bearing and non-weight-bearing activities (such as for example, sitting and standing, walking/jogging and cycling). Validation results obtained in several studies demonstrate applicability to widely varying populations such as healthy individuals and individuals post-stroke, while achieving high (95%-98%) average accuracy of posture and activity classification, high (root-mean-square error of 0.69 METs) accuracy of energy expenditure prediction, and reliable (error of 2.6- 18.6%) identification of temporal gait parameters. High accuracy and minimal intrusiveness of SmartShoe should enable its use in a wide range of research and clinical applications.
Pervasive and Mobile Sensing and Computing for Healthcare Pervasive and Mobile Sensing and Computing for Healthcare Look