Using perceptive computing in multiple sclerosis - the Short Maximum Speed Walk test

Background: we investigated the applicability and feasibility of perceptive computing assisted gait analysis in multiple sclerosis (ms) patients using microsoft kinect™. to detect the maximum walking speed and the degree of spatial sway,we established a computerized and observer-independent measure,which we named short maximum speed walk (smsw),and compared it to established clinical measures of gait disability in ms,namely the expanded disability status scale (edss) and the timed 25-foot walk (t25fw). methods. cross-sectional study of 22 ms patients (age mean ± sd 43 ± 9 years,13 female) and 22 age and gender matched healthy control subjects (hc) (age 37 ± 11 years,13 female). the disability level of each ms patient was graded using the edss (median 3.0,range 0.0-6.0). all subjects then performed the smsw and the timed 25-foot walk (t25fw). the smsw comprised five gait parameters,which together assessed average walking speed and gait stability in different dimensions (left/right,up/down and 3d deviation). results: smsw average walking speed was slower in ms patients (1.6 ± 0.3 m/sec) than in hc (1.8 ± 0.4 m/sec) (p = 0.005) and correlated well with edss (spearman's rho 0.676,p < 0.001). furthermore,smsw revealed higher left/right deviation in ms patients compared to hc. smsw showed high recognition quality and retest-reliability (covariance 0.13 m/sec,icc 0.965,p < 0.001). there was a significant correlation between smsw average walking speed and t25fw (pearson's r = -0.447,p = 0.042). conclusion: our data suggest that ambulation tests using microsoft kinect™ are feasible,well tolerated and can detect clinical gait disturbances in patients with ms. the retest-reliability was on par with the t25fw. © 2014behrens et al.; licensee biomed central ltd.