AI Capable of Diagnosing Autism Using Picture of Eyes

Artificial intelligence isn’t just coming for our jobs.

It’s also expanding rapidly into other fields.

And photography is essential to its mission. 

Not only is artificial intelligence capable of generating images and videos, it can also analyze them. Think space exploration, deciphering long-lost languages, or medicine, as is the case here.

Research scientists revealed that they discovered a way to use photographs of a child’s eyes to assist an AI tool in diagnosing autism in what can only be called a massive leap forward for the field. 

The team at Yonsei University College of Medicine said of their work, “Individuals with ASD have structural retinal changes that potentially reflect brain alterations, including visual pathway abnormalities through embryonic and anatomic connections… Whether deep learning algorithms can aid in objective screening for ASD and symptom severity using retinal photographs is unknown,” Science Alert writes.

Interestingly, the AI was totally accurate when it came to diagnosing autism but wasn’t quite as accurate when it came to gauging the severity of it. To discover this difference, the team let the AI evaluate photographs of children diagnosed with autism and those without it hence the high accuracy on the one measure but the varying output on the second count (48-66% accuracy according to Science Alert).

The research suggests that this could be an integral part of early diagnosis and, over time, perhaps refined to better determine the severity of symptoms.

For our part, it’s nice to read something different involving AI and photography which has dominated our headlines over the past year. After all, people are feeling like this “revolution” is going to put us all out of a job, Adobe included.  

Any thoughts you might have on the future of science and AI as far as photography is concerned are welcome in the comments. 

You can also read some of our other photography headlines at this link.