Scholarlink

Early detection is essential for prompt diagnosis and rehabilitation since abnormal gait patterns are suggestive of conditions like Parkinson’s disease, stroke, and cerebral palsy. Using the Gait Abnormality in Video Dataset (GAVD), An automated computer vision system that examines walking videos to detect irregularities in gait. YOLOv8 is used for human detection, while MediaPipe is used to extract pose keypoints that capture the temporal dynamics of gait. In terms of precision, recall, F1-score, and balanced accuracy, a Transformer encoder outperforms other models and achieves 95.72% accuracy on GAVD when classifying gait into eight categories. The system provides a low-cost, scalable, and non-invasive tool for clinical gait analysis. It is implemented as a real-time Streamlit web app with Twilio-based WhatsApp alerts for aberrant gait notifications.

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

Computer Vision Based Approach for Patient Gait Abnormality Detection and Monitoring

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Bandaru Jaya Nandini,
  • Mamidi Leha Sahithi,
  • Siddareddy Gari Harshika,
  • D. Radha,
  • Nandu C. Nair

摘要

Early detection is essential for prompt diagnosis and rehabilitation since abnormal gait patterns are suggestive of conditions like Parkinson’s disease, stroke, and cerebral palsy. Using the Gait Abnormality in Video Dataset (GAVD), An automated computer vision system that examines walking videos to detect irregularities in gait. YOLOv8 is used for human detection, while MediaPipe is used to extract pose keypoints that capture the temporal dynamics of gait. In terms of precision, recall, F1-score, and balanced accuracy, a Transformer encoder outperforms other models and achieves 95.72% accuracy on GAVD when classifying gait into eight categories. The system provides a low-cost, scalable, and non-invasive tool for clinical gait analysis. It is implemented as a real-time Streamlit web app with Twilio-based WhatsApp alerts for aberrant gait notifications.