Stanford University develops an AI model that determines whether a brain scan image is male or female with over 90% accuracy

It is said that there are differences in the brains of men and women, but until now no clear structural differences have been demonstrated. Dr. Vinod Menon, a neuroscientist at Stanford University, and his colleagues have revealed that they have succeeded in building an AI model that can read brain scan images and determine whether they are male or female with over 90% accuracy. .

AI Determines Sex of Person From Brain Scans - Neuroscience News

Dr. Menon said the main motivation behind his research is that gender differences may play an important role in brain development, aging, and the manifestation of neurological and psychiatric disorders. He says identifying gender differences is an important step in understanding gender-specific issues.

Until now, it has been pointed out that there are differences in brain function between men and women, but no clear structural differences have been found.

As a result of investigating 30 years of data, it was found that there is no significant difference between the brains of men and women - GIGAZINE

“Male brain vs. female brain? ─ Behavior and brain in emotional processing” by Reiko Sawada and Ya Sato of Kyoto University, published in the October 2016 issue of “Psychology World,”

the journal of the Japanese Psychological Association, a public interest incorporated association . In the special feature article titled ``Gender Differences in Human Behavior,'' the article describes an MRI study of the brains of men and women in response to previous research that found that ``the part called the ampulla of the corpus callosum, which is located at the back of the corpus callosum, is larger and bulb-shaped in women than in men.'' It shows an image and shows that it is difficult to distinguish.

Male brain vs female brain? --Sex differences in behavior and brain in emotional processing
(PDF file)

Mr. Menon created a deep neural network model to classify brain image data and performed the task of ``reading a brain scan image of either a man or a woman.'' After repeating the task approximately 1,500 times, the AI model was able to determine with high accuracy whether the scanned image was of a man or a woman.

The points being differentiated are the brain's default mode network, striatum, limbic system network, etc., and the present results suggest that this difference has simply not been detected in the past.

The AI model created by Menon et al. has shown good results on multiple datasets. Furthermore, whether gender differences in the brain are present from early on in life, or whether they are caused by the influence of hormones or differences in social situations is not the subject of this study.

in Science, Posted by logc_nt