Automatic recognition of therapy progress among children with autism

Area of Science:

• Developmental Psychology
• Autism Spectrum Disorder Research
• Human-Computer Interaction

Background:

• Autism Spectrum Disorder (ASD) presents challenges in monitoring therapeutic interventions.
• Objective assessment of therapy progress in children with ASD is crucial for effective treatment.
• Existing methods for progress tracking can be subjective or labor-intensive.

Purpose of the Study:

• To develop and validate a method for recognizing therapy progress in children with ASD.
• To leverage behavioral data from interactive tablet games for objective progress assessment.
• To identify key behavioral parameters that predict therapy outcomes.

Main Methods:

• Children with ASD engaged in five specially designed tablet games.
• Behavioral data were collected via game interactions and tablet sensors (touchscreen, accelerometer, gyroscope).
• Over 180 distinct parameters were calculated from raw data to analyze behavior.

Main Results:

• The proposed method achieved over 80% accuracy in recognizing therapy progress.
• Specific behavioral parameters were identified as strong predictors of progress.
• The study confirmed the feasibility of objective progress recognition in key developmental areas.

Conclusions:

• A novel method using tablet games and data analysis can objectively monitor therapy progress in children with ASD.
• This approach can provide valuable insights for therapists and parents, enabling tailored interventions.
• The findings support the development of a tool to enhance ASD therapy management and reporting.