Scholarlink

<p>In this study, gait anomalies associated with balance disturbances are assessed using metrics describing nonlinear dynamical systems. Three such measures – correlation dimension, largest Lyapunov exponent, and sample entropy – were applied to time series representing 3D rotational angles of skeletal body segments to distinguish between normal and abnormal gait patterns. Motion capture data were collected in a virtual reality environment at the Human Dynamics and Multimodal Interaction Laboratory of the Polish-Japanese Academy of Information Technology, utilizing the Motek CAREN Extended system. Two walking scenarios were analyzed: standard walking and walking across a simulated rope bridge, designed to induce balance-related gait disturbances. Descriptive statistics, including mean and median values, were computed for both scenarios. Statistical analysis confirms significant differences between the two gait types. Furthermore, the performance of anomaly detection is evaluated using Receiver Operating Characteristic – Area Under the Curve (ROC-AUC) and Average Precision (AP) metrics. The results demonstrate the effectiveness of the selected measures, achieving ROC-AUC=0.83 and AP=0.84. These findings support the applicability of nonlinear dynamical metrics in identifying gait anomalies.</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

Gait stability anomalies assessed by measures describing nonlinear dynamical systems

  • 期刊论文

  • 发表日期：

    2026年2月14日

查看全文
  • Adam Świtoński,
  • Henryk Josiński,
  • Tomasz Krzeszowski,
  • Konrad Wojciechowski

摘要

In this study, gait anomalies associated with balance disturbances are assessed using metrics describing nonlinear dynamical systems. Three such measures – correlation dimension, largest Lyapunov exponent, and sample entropy – were applied to time series representing 3D rotational angles of skeletal body segments to distinguish between normal and abnormal gait patterns. Motion capture data were collected in a virtual reality environment at the Human Dynamics and Multimodal Interaction Laboratory of the Polish-Japanese Academy of Information Technology, utilizing the Motek CAREN Extended system. Two walking scenarios were analyzed: standard walking and walking across a simulated rope bridge, designed to induce balance-related gait disturbances. Descriptive statistics, including mean and median values, were computed for both scenarios. Statistical analysis confirms significant differences between the two gait types. Furthermore, the performance of anomaly detection is evaluated using Receiver Operating Characteristic – Area Under the Curve (ROC-AUC) and Average Precision (AP) metrics. The results demonstrate the effectiveness of the selected measures, achieving ROC-AUC=0.83 and AP=0.84. These findings support the applicability of nonlinear dynamical metrics in identifying gait anomalies.