According to research set to be unveiled at the upcoming annual meeting of the Radiological Society of North America (RSNA), a newly developed artificial intelligence (AI) system can effectively diagnose autism in children aged between 24 and 48 months by analyzing specialized MRIs of the brain.
A multidisciplinary team led by B. Mohamed Khudri, which includes Sc., a visiting research scholar at the University of Louisville in Kentucky, developed a three-stage system to analyze and categorize diffusion tensor MRI (DT-MRI) scans of the brain. The team utilizes a unique method called DT-MRI to track the movement of water along light tissue sections in the brain.
Khudri explains that their algorithm is specifically trained to detect variations in brain areas to differentiate between individuals with disabilities and those who are neurotypical.
The AI system, in its quest to evaluate the level of communication among different brain regions, extracts imaging markers from brain tissue images isolated from DT-MRI scans. These imaging markers are then compared between autistic children and typically developing children using a machine learning algorithm.
Dr. Gregory N. Barnes, a co-author and chairman of the Louisville Norton Children’s Autism Center, points out that autism is primarily characterized by atypical connections within the brain. These excessive connections, which manifest as common symptoms in children with autism such as impaired social interaction and repetitive behaviors, are captured through DT-MRI.
The researchers tested their approach using DT-MRI brain images of 226 children aged between 24 and 48 months from the Autism Brain Imaging Data Exchange-II. The dataset comprised scans from 100 typically developing children and 126 children with autism. The technology demonstrated an impressive 97% sensitivity, 98% specificity, and an overall accuracy of 98.5% in identifying children with autism.
Khudri emphasizes that their innovative approach allows for early diagnosis of autism in children under the age of two, enabling prompt medical intervention that could lead to improved outcomes, increased independence, and higher IQ levels for autistic individuals.
The AI system’s report outlines the impacted neural pathways, expected effects on brain function, and a severity rating, guiding early medical interventions based on the identified neural abnormalities.
The experts are currently working towards commercializing their artificial intelligence technology and obtaining FDA approval for its clinical use.
Additional co-authors on the study include Mostafa Abdelrahim, B., Yaser El-Nakieb, Ph.D., Dr. Mohamed Ali, Ph.D., Dr. Ahmed S. ShalabyAli Mahmoud, Ph.D., A. Gebreil, Dr. Ahmed ElnakibAyman S. El-Baz, Ph.D., Andrew Switala, Sohail Contractor, and D. Dr. Ahmed Elnakib.
For more information on the RSNA and access to virtual copies of news releases and digital images from 2023, visit their official website at RSNA.
The RSNA, comprised of oncologists, radiation doctors, medical physicists, and related scientists, is dedicated to advancing patient care and healthcare quality through research and technological advancements. The society is headquartered in Oak Brook, Illinois.
