In the United States approximately 1 in 59 children are born with Autism Spectrum Disorder (ASD). From 2000 to 2014, autism rates increased 250%. Despite the global prevalence of ASD, its etiology remains a mystery lacking a known cure. Signs such as language deficits and impaired social relatedness can be utilized for diagnosis as early as age 2. However, despite these robust detection methods, the average ASD diagnosis does not take place until age 4 or 5.
Gaze aversion, the intentional deflection of attention away from the face the conversation partner, can be used to detect ASD at early ages. While gaze aversion is a robust detection method, the costs of quantitative screenings combined with the need for frequent visits to clinical settings poses significant barriers for such diagnoses.
Given the widespread abundance of smartphones and tablets, as well as the framework and resources to deploy such a study, we propose creating ASDetect: a mobile application for measuring gaze aversion through a series of inputted videos and machine learning algorithms to detect risk for ASD at an early stage. Through an interactive app featuring educational videos and games, we could serve children of any condition and provide early warning signs to parents if gaze aversion is detected.
ASDetect is a platform for eye tracking research, paving the way for future studies in mental health. With a more accurate and faster system, we could potentially detect more conditions, such as Schizophrenia, by tracking the eyes during predictable target movement and detecting adjustments to the gain of pursuit response due to visual and anticipated target movement. Other potential clinical applications for gaze tracking could be to research in OCD, phobias, depression, anxiety, parkinson, optometrics, ocular immunology, and ophthalmic pathology6. Beyond medical research, this technology could be used in human computer interaction, neuromarketing, virtual reality, and gaze interaction for people with physical disabilities.
