Scholarlink

Autism Spectrum Disorder (ASD) involves atypical attention, perception, and language processing. This study examined auditory attention and speech perception in children with ASD using event-related potentials (ERPs) and EEG-based analysis. Short Romanian sentences varying in pitch, distance, direction, and rate were presented while EEG data were recorded with a 16-channel Ultracortex Mark IV system. ERP components (P1, N1, MMN, N2c, P300, N400, LPC) and Power Spectral Density (PSD) features were extracted and analyzed using a Random Forest classifier. Results showed reduced amplitudes in P1, N1, and P300 in ASD, indicating atypical sensory and attentional mechanisms. LPC, N1, and P1 had the highest importance in group differentiation, highlighting deficits in early auditory and higher cognitive processing. These findings suggest that integrating ERP-derived biomarkers with machine learning can support individualized, neurophysiological informed educational and therapeutic programs for children with ASD.

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

A Bridge from Biomedical Engineering to Education: An Event-Related Potential Study Based on Speech Attention in Autism

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • Oana Geman,
  • Sara Sharghilavan,
  • Matti Karppa,
  • Hadi Abbasi,
  • Diana Sinziana Duca,
  • Lucia Morosan-Danila,
  • Cristina Lemni,
  • Tiberiu Ciortan

摘要

Autism Spectrum Disorder (ASD) involves atypical attention, perception, and language processing. This study examined auditory attention and speech perception in children with ASD using event-related potentials (ERPs) and EEG-based analysis. Short Romanian sentences varying in pitch, distance, direction, and rate were presented while EEG data were recorded with a 16-channel Ultracortex Mark IV system. ERP components (P1, N1, MMN, N2c, P300, N400, LPC) and Power Spectral Density (PSD) features were extracted and analyzed using a Random Forest classifier. Results showed reduced amplitudes in P1, N1, and P300 in ASD, indicating atypical sensory and attentional mechanisms. LPC, N1, and P1 had the highest importance in group differentiation, highlighting deficits in early auditory and higher cognitive processing. These findings suggest that integrating ERP-derived biomarkers with machine learning can support individualized, neurophysiological informed educational and therapeutic programs for children with ASD.