[P3-0803] Recovery of walking capacity in stroke rehabilitation using acceleration time-series data analysis：A pilot case study
Keywords:recovery of walking capacity, stroke, acceleration time-series data analysis
【Purpose】
A lot of experience in clinical practice is needed to assess the recovery of walking capacity, with regard to qualitative aspects, based on gait observation. Recently, it was reported that acceleration time-series data was useful to identify gait dynamics which was related to qualitative aspects. The purpose of this study was to investigate whether acceleration time-series data analysis could detect the change with regard to the recovery of walking capacity in patients with stroke.
【Methods】
The participants were two patients with stroke who received physical therapy for four weeks in a rehabilitation care unit. Ten-meter walking test（10MWT）and Dynamic gait index（DGI）were performed to assess the walking capacity. Trunk acceleration was recorded during 10MWT using a tri-axial accelerometer. Using the peak AP accelerations of the non-paralyzed side at heel contact, ten gait cycles were extracted from time-series data. The vertical component in each gait cycle data was divided into seven 64-sample sections with 50% overlapped portions. Within each section, root mean square（RMS）and power spectrum entropy（PSEn）were calculated as parameters representing the magnitude and smoothness of motion, respectively.
【Results】
There were little differences of gait speed between the pre- and post-intervention in both cases. Case A；RMS and PSEn values during heel-contact to terminal-stance phase of theparalyzed leg decreased from the pre- to post-intervention. Case B；RMS values during each section decreased, and PSEn values didn’t change from the pre- to post-intervention. The assessment using RMS and PSEn values in both cases were nearly in accordance with the results from gait observation.
【Discussion】
The results suggested that the acceleration time-series data analysis had a potential value as an assessment tool for the recovery of walking capacity. Further studies using larger sample sizes are needed to assess its value.
A lot of experience in clinical practice is needed to assess the recovery of walking capacity, with regard to qualitative aspects, based on gait observation. Recently, it was reported that acceleration time-series data was useful to identify gait dynamics which was related to qualitative aspects. The purpose of this study was to investigate whether acceleration time-series data analysis could detect the change with regard to the recovery of walking capacity in patients with stroke.
【Methods】
The participants were two patients with stroke who received physical therapy for four weeks in a rehabilitation care unit. Ten-meter walking test（10MWT）and Dynamic gait index（DGI）were performed to assess the walking capacity. Trunk acceleration was recorded during 10MWT using a tri-axial accelerometer. Using the peak AP accelerations of the non-paralyzed side at heel contact, ten gait cycles were extracted from time-series data. The vertical component in each gait cycle data was divided into seven 64-sample sections with 50% overlapped portions. Within each section, root mean square（RMS）and power spectrum entropy（PSEn）were calculated as parameters representing the magnitude and smoothness of motion, respectively.
【Results】
There were little differences of gait speed between the pre- and post-intervention in both cases. Case A；RMS and PSEn values during heel-contact to terminal-stance phase of theparalyzed leg decreased from the pre- to post-intervention. Case B；RMS values during each section decreased, and PSEn values didn’t change from the pre- to post-intervention. The assessment using RMS and PSEn values in both cases were nearly in accordance with the results from gait observation.
【Discussion】
The results suggested that the acceleration time-series data analysis had a potential value as an assessment tool for the recovery of walking capacity. Further studies using larger sample sizes are needed to assess its value.