AI Breakthrough: Toddler autism diagnosis achieves 98.5% accuracy

AI Breakthrough: Toddler autism diagnosis achieves 98.5% accuracy | Credits: Google Image
AI Breakthrough: Toddler autism diagnosis achieves 98.5% accuracy | Credits: Google Image

United States: As artificial intelligence (AI) is making its way towards healthcare, diagnosing autism in toddlers has become easier. According to the latest research conducted by researchers at the University of Louisville, systems developed via AI technology provided near-perfect results in diagnosing the disorder in toddlers.

The accuracy rate of the AI tool reached up to 98.5 percent. The researchers underlined that the technology used specialized MRI scans of the brains to carry out the results.

The findings were presented during a meeting of the Radiological Society of North America (RSNA) in Chicago. However, the reports have claimed that this presented research is preliminary unless it is published in a reviewed journal.

What experts have to say?

Through an RSNA news release, one of the researchers behind the developed system and a visiting research scholar – Mohamed Khudri, stated, “Our algorithm is trained to identify areas of deviation to diagnose whether someone is autistic or neurotypical.”

Visual Representation of Child with Autism Disorder | Credits: Getty

Co-author and professor of neurology and director of the Norton Children’s Autism Centre in Louisville, Dr Gregory Barnes, said, “Autism is primarily a disease of improper connections within the brain,” according to the reports by HealthDay News.

He further said, “DT-MRI captures these abnormal connections that lead to the symptoms that children with autism often have, such as impaired social communication and repetitive behaviors.”

Which technology was used to develop the AI system?

The research has underlined that this AI system was developed on the basis of DT-MRI. The study revealed that the system isolates images from the scans by DT-MRI and surveys the looks at markers to detect the connectivity level between the areas of the brain. Accordingly, the algorithm of the tool compares the patterns in the brains of normal children and children with autism.

According to the researchers, 226 children between the ages of 24 and 48 were examined. Out of the total, 126 brains were affected by autism and the remaining 100 were developing normally. The study revealed that the AI tool had an accuracy of 98.5 percent – as it spotted real cases with 97 percent accuracy and identified the brains without autism with 98 percent.

In this regard, Khudri stated, “Our approach is a novel advancement that enables early detection of autism in infants under two years of age.”

Visual Representation of Artificial Intelligence | Credits: Getty Images

He added, “We believe that therapeutic intervention before the age of three can lead to better outcomes, including the potential for individuals with autism to achieve greater independence and higher IQs,” according to HealthDay News.

US Centres for Disease Control and Prevention recently reported that the developmental evaluation has been received by just half of the children with autism – by the age of fewer than three (3) years. According to the reports by HealthDay News, the study stated around 30 percent of the children with autism were not diagnosed by the age of 8 years.

What do researchers have to say about delayed diagnosis?

While emphasizing the reasons behind the delayed diagnosis, the researchers claimed that one of the reasons could be a lack of resources at the centers. Additionally, Khudri emphasized that this problem can be solved with the help of an AI system – contributing to increased pace.  

Furthermore, Barnes said, “The idea behind early intervention is to take advantage of brain plasticity, or the ability of the brain to normalize function with therapy.”

Besides this, the researchers have stated that they have been looking for approval from the US Food and Drug Administration for the newly developed AI software.

What’s your Reaction?
+1
0
+1
0
+1
0
+1
0
+1
0
+1
0