|
Project Topic
|
Problem: Autism Spectrum Disorder (ASD) is growing (1-in-44 children; 4:1 boys:girls) and manifests in debilitating cognitive problems. “Social blindness”, the inability to recognise emotions in others, is a common debilitating feature, treated via intensive 1:1/small-group therapy. It is costly, in very short supply, and thus often infrequent. Growing ASD diagnosis, particularly in boys, threatens to create a “lost-generation” unable to achieve their full potential.
Need: To prevent severe disorder by significantly increased access to emotional recognition training/therapy for those with ASD, and at low(er) cost.
Proposed Solution: To virtualize emotional recognition therapy by combining three novel key elements (KE1-3)
KE1: Hyper-realistic and responsive avatars able to show detailed emotions (Availability, Scalability, Reproducibility)
KE2: Computer vision to read subject emotional state, reaction rates in therapy tasks, stress levels (via skin impedance, heartrate), executive function (inhibitory control, attention), and behavior patterns, to create critical subject feedback (Personalization)
KE3: Programmed (standard) therapeutic methods to challenge and respond to measured subject response (Adaptive, Gamified)
Combining existing technology creates a fully virtualized, 2-way adaptive, and individualized therapy. The overall solution creates a platform for extension, gamification, and further research based on new insight gains from KE2 data.
Novelty: No such virtualized, interactive, adaptive care exists.
Hypothesis: This two-way technology solution including novel use of subject-specific psycho-physiological response feedback can be =80% as effective as one-to-one therapy with ASD therapists, increasing access to basic care and augmenting/reinforcing current therapy.
Outcome: Highly extensile software-based platform technology solution to dramatically increas-ing access and scalability, lower costs, and create new insights/pathways in ASD research.
|
