Related Concept Videos
You might also read
Related Articles
Articles linked to this work by shared authors, journal, and citation graph.
Sort by
Same author
Integrated precise temperature regulation and electrophysiology sensing system for nanoplasmonic photothermal cardiac bradyarrhythmia therapy.
Microsystems & nanoengineering·2026
Same author
A flow field reconstruction based on cylindrical wake flow data using the PSO-CNN-LSTM algorithm.
Scientific reports·2026
Related Experiment Video
Updated: Jun 18, 2026

04:04
Asthma Detection Research Based on Voice Signal Processing and Machine Learning
Published on: July 22, 2025
Child vocalization composition as discriminant information for automatic autism detection.
Dongxin Xu1, Jill Gilkerson, Jeffrey Richards
1LENA Foundation, Boulder, CO 80301, USA. dongxinxu@lenafoundation.org
Summary
This study introduces an automatic tool for early autism screening in children using speech analysis. The novel system achieves 85-90% accuracy, offering a significant advancement for widespread autism detection.
Area of Science:
- Developmental Psychology
- Speech-Language Pathology
- Machine Learning in Healthcare
Background:
- Early identification of autism spectrum disorder (ASD) is critical for effective early intervention.
- Current autism screening methods rely on parent questionnaires or direct clinical observation, limiting accessibility.
- There is a need for automated, objective tools for widespread childhood autism screening.
Purpose of the Study:
- To develop and validate a fully automatic mechanism for autism detection in young children.
- To leverage speech signal processing and machine learning for objective autism screening.
- To assess the efficacy of the automated tool in differentiating children with autism from language-delayed and typically developing peers.
Main Methods:
- The study extended the Language ENvironment Analysis (LENA) system, utilizing speech signal processing.
- Child vocalization composition data was analyzed using pattern recognition and machine learning algorithms.
- Cross-validation tests were performed on a dataset comprising children with autism, language delays, and typical development.
Main Results:
- Child vocalization composition demonstrated significant discriminant information for autism detection.
- The automated system achieved accuracy rates of 85% to 90% at the equal-error-rate (EER) point.
- The dataset included 34 children with autism, 30 with language delays, and 76 typically developing children.
Conclusions:
- The developed automatic tool shows high accuracy in identifying autism in young children.
- Its ease of use and automatic nature make it suitable for population-based or universal childhood autism screening.
- This technology has the potential to significantly improve early access to intervention services for children with autism.
