A new study by Stanford Medicine investigators has unveiled a new artificial intelligence model that is more than 90 percent successful in determining whether brain activity scans come from a woman or a man.

The results will be published on February 19. Proceedings of the National Academy of Sciences, helps resolve a long-standing controversy over whether reliable sex differences exist in the human brain and suggests that understanding these differences may be important for addressing neuropsychiatric conditions that affect women and men differently. do

“An important motivation for this study is that sex plays an important role in human brain development, aging, and the expression of psychiatric and neurological disorders,” said Vinod Menon, Ph.D., of Psychology and Behavioral Sciences. Professor and director of Stanford said. Cognitive and Systems Neuroscience Laboratory. “Identifying persistent and replicable sex differences in the healthy adult brain is an important step toward a deeper understanding of sex-specific vulnerabilities in psychiatric and neurological disorders.”

Menon is the senior author of the study. The lead authors are senior research scientist Srikanth Riali, Ph.D., and academic staff researcher Yuan Zhang, Ph.D.

The “hotspots” that helped the model most distinguish men’s brains from women’s included the default mode network, a brain system that helps us process self-referential information, and the striatum and limbic net. Work, which is involved in learning and how we respond to rewards. .

The researchers noted that the work did not weigh whether sex-related differences arise early in life or may be due to hormonal differences or different social situations that men and women are more likely to face. May be.

Uncovering mental differences

The extent to which a person’s gender affects how their brain is organized and functions has long been a matter of controversy among scientists. While we know the sex chromosomes we’re born with help determine the cocktail of hormones our brains are exposed to—especially during early development, puberty, and old age—researchers have long have struggled to connect sex with concrete differences in the human brain. Brain structures look similar in men and women, and previous research examining which brain regions work together has largely failed to change consistent brain signals of gender.

In their current study, Menon and his team took advantage of recent advances in artificial intelligence, as well as access to several large data sets, to perform more powerful analyzes than ever before. First, they built a deep neural network model that learns to classify brain imaging data: As the researchers showed the model brain scans and told it was looking at a man’s or a woman’s brain, the model Began to “notice” if subtle patterns could help tell the difference.

The model outperformed previous studies, in part because it used a deep neural network that analyzes dynamic MRI scans. This approach captures the complex interactions between different brain regions. When the researchers tested the model on about 1,500 brain scans, it could almost always tell whether the scan came from a woman or a man.

The success of the model suggests that detectable sex differences in the brain exist but had not been reliably raised before. The fact that it worked so well in a variety of data sets, including brain scans from multiple sites in the US and Europe, makes the results particularly reliable because it controls for many of the confounds that plague these types of studies. can affect

“This is very strong evidence that sex is a strong determinant of the organization of the human brain,” Menon said.

Making predictions

Until recently, a model like Menon’s team would help researchers sort brains into different groups but would not provide information about how the sorting occurred. Today, however, researchers have access to a tool called “descriptive AI,” which can sift through vast amounts of data. please explain How are model decisions made?

Using interpretable AI, Menon and his team identified the brain networks that were most important to the model’s decision whether a brain scan was from a male or a female. They found that the model was looking to the default mode network, the striatum and the limbic network, to make the most frequent calls.

The team then wondered if they could create another model that could predict how well participants would perform on certain cognitive tasks based on functional characteristics of the brain that differed between women and men. . They developed sex-specific models of cognitive abilities: one model effectively predicted cognitive performance in men but not in women, and another in women but not in men. The results suggest that different brain characteristics between the sexes have significant behavioral effects.

“These models worked very well because we successfully separated the brain patterns between the sexes,” Menon said. “This suggests to me that ignoring sex differences in brain organization may be missing the underlying causes of neuropsychiatric disorders.”

While the team applied their deep neural network model to questions about sex differences, Menon says the model could be applied to answer questions about any aspect of brain communication. What type of cognitive ability or behavior might be related? He and his team plan to make their model publicly available for any researcher to use.

“The applicability of our AI models is very broad,” Menon said. which we want to better understand in order to help people adapt and meet these challenges.”

This research was supported by the National Institutes of Health (Grants MH084164, EB022907, MH121069, K25HD074652 and AG072114), The Transdisciplinary Initiative, The Uytengsu-Hamilton 22q11 Research Institute and the UT Ford Institute Health NAAD Institute and the SIGNAL Institute. Sponsored by Tut. Ward