Around one in every 100 children has an autism spectrum disorder (ASD), according to the World Health Organization (WHO), which also points out that, although the characteristic symptoms are usually detected in early childhood, this is not always the case. Furthermore, people with autism may have a disability or suffer from conditions such as epilepsy, depression, anxiety, attention deficit hyperactivity disorder, etc., so early diagnosis and treatment are key to improving their quality of life and that of their loved ones. families.
New research by scientists at the University of South Australia (UniSA) and Flinders University has shown that artificial intelligence (AI) would allow ASD to be diagnosed quickly and accurately with a single flash of light in the eye. Using an electroretinogram (ERG) – a diagnostic test that measures the electrical activity of the retina in response to a light stimulus – researchers have implemented AI to identify specific characteristics to classify ASD.
When they measured the retinal responses of 217 children between 5 and 16 years old (71 with a diagnosis of ASD and 146 children without a diagnosis of ASD), these scientists found that the retina generated a different retinal response in children with ASD compared to those with ASD. They were neurotypical.
The team also found that the most powerful biomarker was achieved with a single flash of bright light in the right eye, and AI processing significantly reduced testing time. The study found that higher frequency components of the retinal signal were reduced in ASD.
Accurately diagnose ASD and related disorders
UniSA researcher Dr Fernando Marmolejo-Ramos says the test could provide doctors with an improved method for diagnosing ASD and speed up care for those affected. “Early interventions and appropriate support can help children with ASD improve their quality of life, but at this time there is no simple ‘test’ for ASD, meaning people often require psychological assessments and reports. prolonged periods to obtain a diagnosis.”
The expert explains that RETeval electroretinogram testing unit can collect data and detect autism in just 10 minutes. He also highlights that “it is non-invasive and children tolerate it well, which makes the process much easier for everyone involved.” The test has been developed in collaboration with the University of Connecticut and University College London, and could be analyzed in depth to confirm whether the results obtained could be used to detect ASD in children and adolescents aged five to 16 years with a high level precision.
“Our study demonstrates the potential of analyzing retinal responses to aid in the identification of neurodevelopmental conditions such as autism spectrum disorder.”
Dr Paul Constable, a researcher at Flinders University and leader of the project, adds that because the eye is connected to the brain, looking inside the eye to understand the brain allows us to learn more about how the brain develops in people with TORCH. “It’s very exciting to start looking at new ways to use electroretinogram with signal analysis and machine learning to help classify ASD more accurately,” he says.
Constable indicates that they have to study younger children, and those with other disorders, such as attention deficit hyperactivity disorder, to determine how specific this test can be, but the findings they have published in JAMA Network They are an important first step to achieve it.
The researchers’ goal is to expand the research to analyze other cohorts and diagnostic categories. “Our study demonstrates the promising potential of analyzing retinal responses using advanced signal processing and machine learning techniques to aid in the identification of neurodevelopmental conditions such as autism spectrum disorder,” says Dr. Hugo Posada-Quintero. , assistant professor in the Department of Biomedical Engineering at the University of Connecticut and co-investigator.
“With further research and technological development, these analytical methods could become practical tools to help clinicians accurately and efficiently detect and diagnose ASD and related disorders,” he concludes.
