Scholarlink

<p>Early Autism Spectrum Disorder (ASD) identification is crucial but resource-intensive. This study evaluated a novel two-stage multimodal AI framework for scalable ASD screening using data from 1242 children (18–48 months). A mobile application collected parent-child interaction audio and screening tool data (MCHAT, SCQ-L, SRS). Stage 1 differentiated typically developing from high-risk/ASD children, integrating MCHAT/SCQ-L text with audio features (AUROC 0.942). Stage 2 distinguished high-risk from ASD children by combining task success data with SRS text (AUROC 0.914, Accuracy 0.852). The model’s predicted risk categories strongly agreed with gold-standard ADOS-2 assessments (79.59% accuracy) and correlated significantly (Pearson <i>r</i> = 0.830, <i>p</i> &lt; 0.001). Leveraging mobile data and deep learning, this framework demonstrates potential for accurate, scalable early ASD screening and risk stratification, supporting timely interventions.</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

Multimodal AI for risk stratification in autism spectrum disorder: integrating voice and screening tools

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2025年8月21日

查看全文
  • Sookyung Bae,
  • Junho Hong,
  • Sungji Ha,
  • Jiwoo Moon,
  • Jaeeun Yu,
  • Hangnyoung Choi,
  • Junghan Lee,
  • Ryemi Do,
  • Hewoen Sim,
  • Hanna Kim,
  • Hyojeong Lim,
  • Min-Hyeon Park,
  • Eunseol Ko,
  • Chan-Mo Yang,
  • Dongho Lee,
  • Heejeong Yoo,
  • Yoojeong Lee,
  • Guiyoung Bong,
  • Johanna Inhyang Kim,
  • Haneul Sung,
  • Hyo-Won Kim,
  • Eunji Jung,
  • Seungwon Chung,
  • Jung-Woo Son,
  • Jae Hyun Yoo,
  • Sekye Jeon,
  • Hwiyoung Kim,
  • Bung-Nyun Kim,
  • Keun-Ah Cheon

摘要

Early Autism Spectrum Disorder (ASD) identification is crucial but resource-intensive. This study evaluated a novel two-stage multimodal AI framework for scalable ASD screening using data from 1242 children (18–48 months). A mobile application collected parent-child interaction audio and screening tool data (MCHAT, SCQ-L, SRS). Stage 1 differentiated typically developing from high-risk/ASD children, integrating MCHAT/SCQ-L text with audio features (AUROC 0.942). Stage 2 distinguished high-risk from ASD children by combining task success data with SRS text (AUROC 0.914, Accuracy 0.852). The model’s predicted risk categories strongly agreed with gold-standard ADOS-2 assessments (79.59% accuracy) and correlated significantly (Pearson r = 0.830, p < 0.001). Leveraging mobile data and deep learning, this framework demonstrates potential for accurate, scalable early ASD screening and risk stratification, supporting timely interventions.