| Autor: |
Naphtaly Ehrenberg, Oluwaseun Ibironke, Akhila Veerubhotla, Rakesh Pilkar |
| Rok vydání: |
2021 |
| Předmět: |
|
| Zdroj: |
Archives of Physical Medicine and Rehabilitation. 102:e79 |
| ISSN: |
0003-9993 |
| Popis: |
Research Objectives To compare the accuracy of physical activity monitor-based step-count algorithm to determine the best wear location in individuals with hemiparetic gait. Design Cross-sectional study- two lab visits which are one week apart. Setting Motion Analysis lab. Participants Three individuals with hemiparesis post-stroke, aged 62.3±2.1 years. Interventions Participants completed a 10-meter walk test, five 25-meter walk tests, two 2-minute periods and a 6-minute period of continuous walking on two different lab visits that were one week apart. Participants wore three ActiGraph GT9X Link activity monitors, one each on their unaffected ankle, waist, and wrist respectively. All walking tests were video-recorded and the criterion measure of step count for each test was obtained through visual counting. A custom peak-detection algorithm built-in Matlab software was used to detect the step count from the ActiGraph activity monitors. The step counts detected from the ActiGraph monitor were compared against the criterion (video-based visual) step counts for all tests on both lab visits. Main Outcome Measures Step count. Results The step count detection algorithm was more accurate at the ankle and waist compared to the wrist. The algorithm had an accuracy that ranged from 85% to 100% across the walking tests. It was observed that the algorithm had greater accuracy for walking bouts that lasted longer and involved more steps (>10 steps) compared to walking bouts that contained fewer steps (∼ 6-7 steps). Conclusions Ankle and waist-worn activity monitors may provide better accuracy in detecting step counts compared to wrist-worn monitors in individuals with hemiplegic gait. Author(s) Disclosures No bias or conflict to disclose. |
| Databáze: |
OpenAIRE |
| Externí odkaz: |
|
