Scholarlink

Background <p>Wearable sensor technology has revolutionized gait analysis by enabling studies in real-world settings and enhancing ecological validity. However, this shift introduces challenges due to uncontrolled environmental factors that contribute to variability, reflecting the multiple contexts in which real-world gait takes place. This study investigates the variability of gait acceleration data, collected with wearable sensors in remote settings, compared to controlled laboratory measurements, focusing on within-individual variability over time.</p> Methods <p>Data were collected from two groups: ten healthy individuals who wore SENS Motion activity monitors on their thighs continuously for two weeks during daily activities, and fifty participants from an open-access gait laboratory dataset. The remote data enabled analysis of gait in natural, unsupervised settings, while the laboratory data provided a controlled comparison. Variability in acceleration time profiles across strides was analyzed using mixed-effect models, and dynamic stability was assessed through the largest Lyapunov exponent (LLE) to evaluate gait consistency and stability.</p> Results <p>The results revealed that acceleration data from the remote settings exhibited higher variability both within and between days compared to the controlled laboratory data. This increased variability was most pronounced along the anteroposterior axis, likely due to the uncontrolled nature of the remote environment. Despite this variability, the dynamic stability of the gait patterns, as measured by LLE, showed no significant differences between the two environments; however, not comparing the same individuals in both settings may have obscured potential differences in LLE.</p> Conclusions <p>The findings suggest that while wearable sensors capture greater variability in uncontrolled environments, the underlying stability of gait patterns remains intact. The perservation of stability despite greater variability supports the use of wearable sensors for extended gait monitoring in remote settings, with potential implications for improving remote patient monitoring and understanding gait variability in clinical populations.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Assessing within-individual gait acceleration variability: a comparative analysis of remote monitoring and laboratory measurements

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2025年9月23日

查看全文
  • Arash Ghaffari,
  • Reed D. Gurchiek,
  • Rasmus Waagepetersen,
  • John Rasmussen,
  • Søren Kold,
  • Ole Rahbek

摘要

Background

Wearable sensor technology has revolutionized gait analysis by enabling studies in real-world settings and enhancing ecological validity. However, this shift introduces challenges due to uncontrolled environmental factors that contribute to variability, reflecting the multiple contexts in which real-world gait takes place. This study investigates the variability of gait acceleration data, collected with wearable sensors in remote settings, compared to controlled laboratory measurements, focusing on within-individual variability over time.

Methods

Data were collected from two groups: ten healthy individuals who wore SENS Motion activity monitors on their thighs continuously for two weeks during daily activities, and fifty participants from an open-access gait laboratory dataset. The remote data enabled analysis of gait in natural, unsupervised settings, while the laboratory data provided a controlled comparison. Variability in acceleration time profiles across strides was analyzed using mixed-effect models, and dynamic stability was assessed through the largest Lyapunov exponent (LLE) to evaluate gait consistency and stability.

Results

The results revealed that acceleration data from the remote settings exhibited higher variability both within and between days compared to the controlled laboratory data. This increased variability was most pronounced along the anteroposterior axis, likely due to the uncontrolled nature of the remote environment. Despite this variability, the dynamic stability of the gait patterns, as measured by LLE, showed no significant differences between the two environments; however, not comparing the same individuals in both settings may have obscured potential differences in LLE.

Conclusions

The findings suggest that while wearable sensors capture greater variability in uncontrolled environments, the underlying stability of gait patterns remains intact. The perservation of stability despite greater variability supports the use of wearable sensors for extended gait monitoring in remote settings, with potential implications for improving remote patient monitoring and understanding gait variability in clinical populations.