Scholarlink

<p>Autism Spectrum Disorder is characterized by multiple social and behavioral challenges. One of the most prominent symptoms is the potential aggressiveness in autistic children, which can sometimes lead to physical harm. This research introduces an innovative Digital Twin-based monitoring framework aimed at predicting irregularities in the physical activity of autistic children to address these behavioral issues. The main goal is to enhance indoor safety measures for children by generating timely alerts for caregivers and healthcare professionals. The study proposes an activity assessment algorithm over a digital twin platform that combines the Bayesian technique with hybrid Convolutional Neural Networks to recognize physical anomalies. Additionally, the proposed framework ensures data security through advanced blockchain technology. It employs the Reputation-based Byzantine Fault Tolerance method for consortium networks within the blockchain. The effectiveness of this approach is assessed against leading research techniques using a comprehensive dataset of 90,367 instances. The results demonstrate significant improvements in key performance metrics, including Temporal Efficacy (10.4 seconds), Classification Efficacy (Precision 96.49%, Sensitivity 95.59%, Specificity 94.34%, F-Measure 94.49%), Decision-Making Efficiency (<InlineEquation ID="IEq1"> <EquationSource Format="TEX">\(r^2 = 96\%\)</EquationSource> </InlineEquation>, Error Rate 0.34%), and Stability (78%).</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

Digital Twin framework for Autism Spectrum Disorder analysis

  • 期刊论文

  • 发表日期：

    2026年7月29日

查看全文
  • Abdullah Alqahtani,
  • Munish Bhatia

摘要

Autism Spectrum Disorder is characterized by multiple social and behavioral challenges. One of the most prominent symptoms is the potential aggressiveness in autistic children, which can sometimes lead to physical harm. This research introduces an innovative Digital Twin-based monitoring framework aimed at predicting irregularities in the physical activity of autistic children to address these behavioral issues. The main goal is to enhance indoor safety measures for children by generating timely alerts for caregivers and healthcare professionals. The study proposes an activity assessment algorithm over a digital twin platform that combines the Bayesian technique with hybrid Convolutional Neural Networks to recognize physical anomalies. Additionally, the proposed framework ensures data security through advanced blockchain technology. It employs the Reputation-based Byzantine Fault Tolerance method for consortium networks within the blockchain. The effectiveness of this approach is assessed against leading research techniques using a comprehensive dataset of 90,367 instances. The results demonstrate significant improvements in key performance metrics, including Temporal Efficacy (10.4 seconds), Classification Efficacy (Precision 96.49%, Sensitivity 95.59%, Specificity 94.34%, F-Measure 94.49%), Decision-Making Efficiency ( \(r^2 = 96\%\) , Error Rate 0.34%), and Stability (78%).