Artificial intelligence (AI) technology is being proposed as a tool to aid in the diagnosis of autism spectrum disorder. A recent article in Scientific Reports highlights the efforts of researchers from Brazil, France, and Germany who employed Magnetic Resonance Imaging (MRI) to train a machine learning algorithm for this purpose.
The researchers put forth a “quantitative diagnostic method” based on brain imaging data collected from 500 individuals, including over 240 diagnosed with autism. Utilizing machine learning techniques, the team embarked on developing their methodology by gathering functional MRI (fMRI) and electroencephalogram (EEG) data. Comparing maps of individuals with and without autism spectrum disorder, the researchers identified the potential for diagnosis through this method.
The machine learning algorithm was then trained using these maps, enabling the system to achieve an accuracy rate exceeding 95% in determining the brain alterations associated with autism.
Contrasting previous research that leans on machine learning for autism diagnosis, this study emphasizes brain network organization rather than relying on a single statistical parameter. Examination of the fMRI data unveiled modifications in specific brain regions linked to cognitive, emotional, learning, and memory processes. Moreover, cortical networks in autism patients exhibited higher segregation, reduced information distribution, and diminished connectivity compared to control subjects.
Francisco Rodrigues, the last author of the article and a professor at the University of São Paulo’s Institute of Mathematics and Computer Science, commented, “Until a few years ago, little was known about the alterations that lead to the symptoms of ASD. Now, however, brain alterations in ASD patients are known to be associated with certain behaviors, although anatomical research shows that the alterations are hard to see, making diagnosis of mild ASD much harder. Our study is an important step in the development of novel methodologies that can help us obtain a deeper understanding of this neurodivergence.”
It’s important to note that this methodology is still in its developmental stage and will likely take years to fully implement, as stated by the São Paulo Research Foundation, which supported the research.
According to the Centers for Disease Control and Prevention, approximately one in 36 children has been identified with autism spectrum disorder. Diagnosis of this developmental disability is challenging due to the lack of a definitive medical test, such as a blood test, for diagnosis.
