A Multimodal Data-Driven Assessment System for Autism Spectrum Disorder in Children: Development and Pilot Validation...

Area of Science:

• Neuroscience
• Biomedical Engineering
• Developmental Psychology

Background:

• Autism Spectrum Disorder (ASD) diagnosis relies on subjective tools like ADOS and CARS, which have limitations in accuracy and efficiency.
• Early and accurate ASD screening is crucial for timely intervention and improved outcomes.
• Existing diagnostic methods can be time-consuming and may not capture the full spectrum of ASD characteristics.

Purpose of the Study:

• To develop and validate a portable, multi-modal data acquisition platform for early Autism Spectrum Disorder (ASD) screening.
• To integrate various data streams including EEG, ECG, speech, and facial expressions for comprehensive ASD assessment.
• To evaluate the platform's effectiveness in identifying distinguishing features between ASD and control groups.

Main Methods:

• Development of a portable multi-modal data acquisition system integrating EEG, ECG, speech, facial expressions, and rating-scale data.
• Utilized an algorithmic framework for data fusion and analysis to generate individualized diagnostic reports.
• Conducted a pilot study with seven participants to validate the system's effectiveness in early ASD screening.

Main Results:

• The multi-modal platform identified significant differences between ASD and control groups.
• Key findings include shorter speech pause duration (41.8% reduction, p < 0.001), increased EEG δ-band power (226% increase, p = 0.0015), and decreased approximate entropy (38.5% reduction, p < 0.0001).
• Pilot data supports the effectiveness of multi-modal data fusion for ASD assessment.

Conclusions:

• The developed multi-modal ASD assessment system demonstrates advantages in low cost, high accuracy, and user-friendliness compared to conventional tools.
• The system holds strong potential for widespread application in family-based ASD screening and early intervention.
• Further data expansion and technical optimization are expected to enhance its clinical utility and diagnostic accuracy for ASD.