The layer facilitates communication between different regions of the brain. Men and women are known to have different rates and symptoms of multiple sclerosis, autism spectrum disorder, migraines, and other brain issues. Understanding how biological sex impacts the brain can improve diagnostic tools and treatments. However, researchers only have a partial understanding of the brain’s cellular layout, despite exploring brain size, shape, and weight. A new study by researchers at NYU Langone Health used machine learning, an AI technique, to analyze.The study analyzed thousands of MRI brain scans from 471 men and 560 women. The results showed that computer programs were able to effectively differentiate between male and female brains based on patterns in structure and complexity that were undetectable to the human eye. This discovery was confirmed by three different AI models specializing in identifying biological sex, each with unique strengths in focusing on specific areas of white matter or examining relationships across larger brain regions.
“Our findings offer a better understanding of the structure of the living human brain, which could lead to new insights into psychological processes.””Iatric and neurological disorders develop and manifest differently in men and women,” explained Yvonne Lui, MD, the senior author of the study and a neuroradiologist. Lui, who is also a professor and vice chair for research in the Department of Radiology at NYU Grossman School of Medicine, pointed out that previous research on brain microstructure has mainly been based on animal models and human tissue samples. Furthermore, the accuracy of some of these previous findings has been questioned due to their reliance on statistical analyses of “hand-drawn” regions of interest, which required subjective decisions about their shape, size, and location.The results of the study, which will be published online on May 14 in the journal Scientific Reports, could potentially be influenced by the regions selected for analysis, according to Lui. To avoid this issue, the researchers utilized machine learning to analyze entire groups of images without pinpointing specific areas, thereby eliminating human biases. To conduct the study, the team provided AI programs with existing brain scan data from both healthy men and women, as well as the biological sex of each scan. These models were specifically designed to utilize complex statistical and algorithmic methods.Mathematical techniques improved with more data and learned to distinguish biological sex without using brain size and shape, according to Lui. The models correctly identified the subject’s sex between 92% and 98% of the time, with water movement through brain tissue being a crucial determining factor. These results emphasize the significance of delving into the brain’s intricate anatomical features.”Studying diseases that affect the human brain requires a diverse approach that includes considering sex differences,” said Junbo Chen, MS, a doctoral candidate at NYU Tandon School of Engineering, who co-led the study.
“Historically, using men as the standard model for various disorders may lead researchers to overlook important insights,” added Vara Lakshmi Bayanagari, MS, a graduate research assistant at NYU Tandon School of Engineering, also a co-lead author of the study.
Bayanagari warns that while AI tools can identify differences in brain-cell organization, they cannot determine which sex may be more likely to have certain features. The study categorized sex as a binary variable, but the authors caution that sex is a spectrum.The study was funded by grants from the National Institutes of Health and the United States Department of Defense. The research team plans to further investigate the development of sex-related brain structure differences over time, taking into account environmental, hormonal, and social factors. Other researchers from NYU Langone Health and NYU were also involved in the study along with Lui, Chen, and Bayanagari.Chung, PhD, and Yao Wang, PhD.
Â
