Autism Genetic Markers Deciphered with AI, Glimmers of Hope for Diagnosis

Diagnose with certainty and treat autism – and related neurological conditions – thanks to artificial intelligence that innovatively re-elaborates images of brain MRI. This is the hope for the future, which comes from research published in ‘Science Advances’ and coordinated by the University of Virginia in the USA. The multidisciplinary team – also made up of biomedical and computer engineers – led by Professor Gustavo K. Rohde has developed a system capable of identifying the genetic markers of autism “in images of brain MRI with an accuracy of 89-95%”, the research highlights.

“Autism is traditionally diagnosed based on the behavior of the patient – the authors recall – but it has a solid genetic basis. A DNA-based approach “could transform the understanding and treatment of the disease”. The basis of the research are the so-called ‘duplications’, or some repeated genetic sequences; those diagnosed with autism have a greater or lesser presence of these ‘duplications’. The researchers have developed a new approach called Tbm (or ‘transport-based morphometry’) that allows to distinguish the normal biological variations in the brain structure from those associated with ‘duplications’. This process has been examined by artificial intelligence associated with brain magnetic resonance imaging. In short, the new approach allows to overcome the obstacles that until now have prevented understanding the ‘gene-brain-behavior’ relationship linked to the development of autism, effectively limiting specialists to diagnoses or treatments based only on behavior.

For their study, the team used data from participants in the Simons Variation in Individuals Project, a group of people with the genetic variation linked to autism. “We hope that the results, the ability to identify localized changes in brain morphology linked to changes in the number of ‘duplications,’ can point to brain regions and ultimately mechanisms linked to autism that can be exploited for therapies,” Rohde said.