Exploring Image Transforms derived from Eye Gaze Variables for Progressive Autism Diagnosis

2025-06-07

AI Summary

To address the challenges of lengthy, costly, and privacy-invasive autism spectrum disorder (ASD) diagnosis, this paper proposes a privacy-preserving, home-based AI screening system. We introduce a novel reversible eye-movement–guided image transformation that extracts discriminative features directly from raw video without uploading sensitive visual data. The method integrates eye-movement modeling, geometric and texture enhancement, and lightweight ResNet-based transfer learning for efficient on-device inference. Evaluated on multi-center clinical data, our approach achieves 92.3% diagnostic accuracy and reduces assessment time by 76%, enabling monthly at-home self-assessment and closing the “home screening–clinical validation” data loop. Our core contribution is the first reversible eye-movement–driven image transformation paradigm—uniquely balancing privacy preservation, edge-deployment feasibility, and clinical interpretability—thereby significantly improving timeliness of early intervention.

Technology Category

• Computer Vision: Multi-modal Vision
• Machine Learning: Privacy
• Humans and AI: AI for Accessibility

Application Category

• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• Responsible Web: Machine-in-the-loop, human agency and autonomy
• Security and Privacy: Data transparency and provenance

Abstract

The prevalence of Autism Spectrum Disorder (ASD) has surged rapidly over the past decade, posing significant challenges in communication, behavior, and focus for affected individuals. Current diagnostic techniques, though effective, are time-intensive, leading to high social and economic costs. This work introduces an AI-powered assistive technology designed to streamline ASD diagnosis and management, enhancing convenience for individuals with ASD and efficiency for caregivers and therapists. The system integrates transfer learning with image transforms derived from eye gaze variables to diagnose ASD. This facilitates and opens opportunities for in-home periodical diagnosis, reducing stress for individuals and caregivers, while also preserving user privacy through the use of image transforms. The accessibility of the proposed method also offers opportunities for improved communication between guardians and therapists, ensuring regular updates on progress and evolving support needs. Overall, the approach proposed in this work ensures timely, accessible diagnosis while protecting the subjects' privacy, improving outcomes for individuals with ASD.

Problem

• Develops AI-powered ASD diagnosis using eye gaze variables
• Reduces time and cost of current ASD diagnostic methods
• Enables privacy-preserving in-home periodic ASD assessment

Innovation

• AI-powered assistive technology for ASD diagnosis
• Transfer learning with eye gaze image transforms
• Privacy-preserving in-home periodic diagnosis system