Quantification of Walking Independence Measure in Hemiplegic Patients using Wearable Motion Sensor

DOI Web Site

Bibliographic Information

Other Title
  • ウェアラブルモーションセンサを用いた片麻痺患者における歩行自立度の定量化
  • ウェアラブルモーションセンサ オ モチイタ ヘンマヒ カンジャ ニ オケル ホコウ ジリツド ノ テイリョウカ

Search this article

Abstract

In this study, we attempted to develop a wearable motion measurement system and to propose a quantitative judgment method of walking independence in poststroke hemiplegic patients based on information of relative angles. The system was consisted of three sensor units with a tri-axial accelerometer and three gyroscopes, a transmitter, a buttery and a personal computer with a receiver. The sensor units were fixed to the subject's waist and both of thighs, respectively. 35 hemiplegic patients participated and performed 10 m walking test. 20 subjects were able to walk independently and the others could only walk under supervision. As criterion indexes for judgment of walking independence, walking cycle, ratio of extension-flexion time, excursion, their coefficients of variance, regularity of angle patterns, dependency between waist and thigh angles were calculated from information of relative angles. The analysis showed that it was difficult to classify the independent group and the supervised group with each criterion index. In almost all supervised subjects, however, either of those criterion indexes deviated from the standard range of independent group. Our proposed quantitative judgment method of walking independence consisted of eight criterion indexes, which were selected by their test sensitivities and positive predictive values. The subject was judged to be a walk under supervision when two criterion indexes or more were positive. As a result, the proposed method classified 33 subjects correctly. Our results suggest that the proposed system and judgment method of walking independence can be very useful to assess the independence measure for walking quantitatively.

Journal

Related Projects

See more

Keywords

Details 詳細情報について

Report a problem

Back to top