Autism diagnosis among girls has historically proven more difficult compared to boys, but new research, which has identified different patterns in brain organization between males and females with the condition, could change this. When researchers set artificial intelligence (AI) to the task of categorizing brain scans between male and female, it was able to filter out the female films with 86 percent accuracy.

The study, published in The British Journal of Psychiatry, believes that understanding the way brain organization differs between males and females with autism could pave the way for better screening for girls who are currently harder to diagnose. This is because it could highlight differences in the symptomatic presentation between sexes, making female autism easier to spot.

“We detected significant differences between the brains of boys and girls with autism, and obtained individualized predictions of clinical symptoms in girls,” said senior author and professor of psychiatry Dr Vinod Menon in a statement. “We know that camouflaging of symptoms is a major challenge in the diagnosis of autism in girls, resulting in diagnostic and treatment delays.”

First, the team gathered functional magnetic resonance imaging brain scans for 773 children with autism. Of those, 637 were boys and 136 girls. That they couldn’t get equal numbers for both sexes both reflects the disparity in the diagnosis of, and research into, girls with autism, and it also complicated their study

Looking for differences in sets of data using AI is usually done using sets of data that are roughly equal. Luckily, the team were able to utilize a new method devised by co-author and assistant professor of computer science and of statistics at Stanford Dr Tengyu Ma which could sort between unequal sets of data.

Running the algorithm revealed that AI could distinguish between boys and girls with 86 percent accuracy. They were also able to establish that the organization differences were autism-dependent after running 976 brain scans of children without autism which the algorithm couldn't tell apart.

The research has the potential to improve the diagnosis of autism in females in addressing that the condition presents in the brain differently and therefore likely exhibits different behavioral symptoms. Taking this into account could one day enable doctors to diagnose autism among girls and boys more equally as it doesn’t only take male-associated symptoms into account.

“When a condition is described in a biased way, the diagnostic methods are biased,” said lead author Dr Kaustubh Supekar in a statement. “This study suggests we need to think differently.”