The University of Louisville in Kentucky, USA, has introduced an innovative artificial intelligence (AI) system that can accurately diagnose autism in children aged 24 to 48 months with an impressive accuracy rate of 98.5%. Led by Mohamed Khudri, a visiting research scholar, the AI system utilizes a three-stage process that focuses on analyzing diffusion tensor MRI (DT-MRI) scans of the brain. By identifying areas of deviation, the algorithm in the AI system can diagnose whether someone is autistic or neurotypical.
To validate the AI system, it was rigorously evaluated using DT-MRI scans from 226 children aged 24 to 48 months, including 126 diagnosed with autism and 100 typically developing children. The results were remarkable, with the AI system demonstrating a sensitivity of 97%, specificity of 98%, and an overall accuracy rate of 98.5% in identifying children with autism.
One of the significant advantages of this AI system is its potential for early detection of autism in infants under two years of age. This early detection could lead to therapeutic intervention before the age of three, resulting in improved outcomes such as greater independence and higher IQs for individuals with autism.
The urgency for early intervention is highlighted by the fact that less than half of children with autism spectrum disorder undergo a developmental evaluation by the age of three, and 30% of those meeting the criteria for autism spectrum disorder do not receive a formal diagnosis by the age of eight. The AI system has the potential to address this issue by streamlining the assessment process, starting with DT-MRI scans followed by a confirmation session with a psychologist to guide parents on the next steps.
In addition to its diagnostic capabilities, the AI system generates a comprehensive report that details the specific neural pathways affected, the expected impact on brain functionality, and a severity grade. This information is crucial in guiding early therapeutic intervention and tailoring treatments to the unique needs of each child.
The University of Louisville’s research team is actively working towards commercializing and obtaining FDA clearance for their AI software. This step towards making the technology widely accessible marks a significant milestone in translating groundbreaking research into real-world applications.
The introduction of the University of Louisville’s AI system represents a paradigm shift in the field of autism diagnosis. With its exceptional accuracy and potential for early intervention, this technology has the power to transform the lives of individuals with autism and revolutionize healthcare delivery. The unveiling of this achievement at the Radiological Society of North America (RSNA) further emphasizes the crucial role of AI in advancing medicine and our understanding and treatment of neurodevelopmental disorders.