Scholarlink

<p>Interactive physiotherapy and rehabilitation systems require accurate motion assessment and timely feedback to support effective training and user engagement. This study proposes an integrated Human–Computer Interaction (HCI) framework combined with LSTM networks for analyzing sequential 3D joint-trajectory data and providing interactive feedback for motion assessment. The proposed pipeline includes motion data preprocessing and normalization, temporal sequence modeling using LSTM to capture movement dynamics, and an interactive interface for delivering feedback in near real time. We evaluate the framework on motion sequences derived from the Human3.6&#xa0;M dataset under the described experimental setting. The proposed method achieves up to 91% classification accuracy, reduces average data processing time to 2.1&#xa0;s, and lowers measured computational resource usage by 50–65% compared with the implemented baselines under the same protocol. Usability is evaluated using the System Usability Scale (SUS), and the detailed usability outcomes are reported in the usability evaluation section. These results support the feasibility of combining interactive interfaces with recurrent sequence modeling for physiotherapy motion assessment, while broader validation on physiotherapy-specific patient datasets remains an important direction for future work.</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

Design and evaluation of an interactive physiotherapy motion-assessment system using HCI and LSTM networks

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年5月20日

查看全文
  • Nahideh DerakhshanFard,
  • Alireza Koshgizadeh,
  • Amir Mollanejad,
  • Abbas mirzaei,
  • Rasoul Farahi

摘要

Interactive physiotherapy and rehabilitation systems require accurate motion assessment and timely feedback to support effective training and user engagement. This study proposes an integrated Human–Computer Interaction (HCI) framework combined with LSTM networks for analyzing sequential 3D joint-trajectory data and providing interactive feedback for motion assessment. The proposed pipeline includes motion data preprocessing and normalization, temporal sequence modeling using LSTM to capture movement dynamics, and an interactive interface for delivering feedback in near real time. We evaluate the framework on motion sequences derived from the Human3.6 M dataset under the described experimental setting. The proposed method achieves up to 91% classification accuracy, reduces average data processing time to 2.1 s, and lowers measured computational resource usage by 50–65% compared with the implemented baselines under the same protocol. Usability is evaluated using the System Usability Scale (SUS), and the detailed usability outcomes are reported in the usability evaluation section. These results support the feasibility of combining interactive interfaces with recurrent sequence modeling for physiotherapy motion assessment, while broader validation on physiotherapy-specific patient datasets remains an important direction for future work.