AI eye test for autism in kids may be able to tell how severe it is

Embargoed until: Publicly released: 2023-12-16 03:00

International

CC-0. https://pixabay.com/photos/boy-brown-eyes-eyes-face-male-4948961/

Photographs of children's retinas - the 'screen' at the back of the eye - can be used to test for autism, and possibly even for the severity of the disease, with a little help from artificial intelligence (AI), according to Korean scientists. The team tested the AI on images of the retinas of 958 children with an average age of 7.8 years, including a total of 1,890 images of individual eyes. Half the kids had an autism diagnosis, while the other half did not. The AI was able to pick out those with an autism diagnosis perfectly based on the images, and determine the severity of symptoms correctly between 48% and 66% of the time. Eye tests could speed up the process of diagnosing autism in young children, the authors say, taking the pressure off specialised child psychiatrists who currently diagnose the condition.

Media release

From: JAMA

Development of Deep Ensembles to Screen for Autism and Symptom Severity Using Retinal Photographs

About The Study: The findings of this diagnostic study of 1,890 eyes of 958 participants support the potential of artificial intelligence as an objective tool in screening for autism spectrum disorder (ASD) and possibly for symptom severity using retinal photographs. Retinal photograph use may speed the ASD screening process, which may help improve accessibility to specialized child psychiatry assessments currently strained by limited resources.

Attachments

Note: Not all attachments are visible to the general public. Research URLs will go live after the embargo ends.

Research JAMA, Web page The URL will go live after the embargo ends

Journal/
conference: JAMA Network Open

Research:Paper

Organisation/s: Yonsei University College of Medicine, Republic of Korea

Funder: This research was supported by faculty research grant 6-2020-0232 from Yonsei University College of Medicine (Prof Cheon), grant 2021R1A2C2010913 from the Korean government to the Basic Science Research Program through the National Research Foundation of Korea, grant MHER22A01 from the National Center for Mental Health (Prof Cheon), grant DUCR-000050 from the Student Research Bursary of Yonsei University College of Medicine, and a grant from the MD-PhD/Medical Scientist Training Program through the Korea Health Industry Development Institute, funded by the Ministry of Health andWelfare, Republic of Korea.

Media Contact/s

Contact details are only visible to registered journalists.