New App that Tracks Eye Movement to Diagnose Autism Could Fast

The IEEE Wireless Health Conference  2016 saw the possibility of an app that can detect the presence of Autism symptoms among children as young as 2 years old.

A State University of New York, Buffalo student, Kun Woo Cho, has such a smartphone app in mind. She is a research assistant in the department of Computer Science and Engineering.

Her research focuses primarily on biometrics based personal identification and security.

This app aims to detect the presence of autism by calculating eye movement, thus allowing faster detection compared to repeated visits to a doctor or a carer.

 

Early detection is still considered the best way to counter the intensifying symptoms of Autism and ADHD, which make social life almost impossible for people in the spectrum.

Behavioural therapy can be started anytime during the lifetime of a autistic person, however, therapy from a very young age can counter and deter the symptoms and allow better development.

Michelle Hartley-McAndrew (MD – clinical assistant professor at Jacobs School in the Department of Paediatrics, Neurology and Biomedical Sciences), who co-authored this study said,

[su_quote]The brain continues to grow and develop after birth. The earlier the diagnosis, the better, . . . Then we can inform families and begin therapies which will improve symptoms and outcome.[/su_quote]

According to Cho,

[su_quote]This is an ongoing study on how to analyze ASD by monitoring gaze patterns. I used the Wasserstein metric, designed the system protocol, and visual stimuli using social scenes.[/su_quote]

Cho is working under the direction of her advisor Wenyao Xu at the Embedded Sensing and Computing (ESC) Lab.

The eye movement of autistic people vary from normal people  and Cho intends to use this to detect autism traits in children.

The application, conceptually termed as, the Early Screening Approach to Autism Spectrum Using Discrete Gaze Pattern, in Kun Cho’s research, details it as a new method developed to screen ASDs, which uses ‘1st Wasserstein distance for dissimilarity measure between gaze distributions’

The pilot study had included 32 children, aged 2-10 years, and investigated the impact of the visual stimuli types on the performance of Gaze-Wasserstein.

The app was found to have an accuracy of 93.96 percent.

According to Cho, the app is still a prototype.

[su_quote]We have to consider if other neurological conditions are included, like ADD, how that will affect the outcome.[/su_quote]

The app merely takes a minute, almost 54 seconds to detect gaze patterns, which makes it a non-intrusive procedure. The short time span is also a boon, when it comes to the extremely short attention span of children with autism and ADHD.

It will be easy to use, given the smartphone format, whereupon, a parent can use it by themselves saving them a lot of valuable time and head straight to therapy.

It is now a known fact that families dealing with autism do not have as much help as necessary, given the waitlist and the time consuming diagnosis.

Further research is in process that involves a much larger sample size of 300 to 400 children, a number almost equal to the annual enrollment for new evaluations at Autism Spectrum Disorder Centre, Buffalo.

According to some of Cho’s colleagues, this project might be the one that can eliminate the diagnosis waitlist and take families straight to treatment.

As of now, though a few Autism tests are available online, there is no full-proof method to detect autism. All the diagnosis is based on opinions and experience of experts, and therefore, the range of intensity varies greatly, without anything to objectively measure the symptoms.

An app that uses gaze pattern can do just that – an objective evaluation on the intensity of the symptoms, channelling more people into the right kind and degree of therapy necessary for them.