Abstract: Researchers use AI to reveal cellular-level differences in the brains of men and women, focusing on white matter. These findings suggest that AI can accurately identify sex-based brain patterns hidden from the human eye.
The study suggests that understanding these differences could improve diagnostic tools and treatments for mental disorders. This research emphasizes the need for diversity in brain studies to ensure a comprehensive understanding of neurodegenerative diseases.
Important facts:
- AI accuracy: AI models identify biological sex in MRI scans with 92%-98% accuracy.
- White Matter Focus: Differences were found in the brain's white matter, which is important for interregional communication.
- Better diagnostics: Understanding gender-based brain differences could improve diagnosis and treatment of diseases such as multiple sclerosis and autism.
Source: NYU Langone
Artificial intelligence (AI) computer programs that process MRI results reveal differences in how men's and women's brains are organized at the cellular level, a new study suggests. These changes were seen in white matter, the tissue primarily located in the innermost layer of the human brain, which promotes communication between regions.
Men and women are known to experience multiple sclerosis, autism spectrum disorder, migraines, and other brain problems at different rates and with different symptoms.
A detailed understanding of how biological sex affects the brain is seen as a way to improve diagnostic tools and treatments.
However, while brain size, shape, and weight have been explored, researchers have only a partial picture of how the brain is organized at the cellular level.
Led by researchers at NYU Langone Health, the new study used an AI technique called machine learning to analyze thousands of MRI brain scans of 471 men and 560 women.
The results show that computer programs can accurately distinguish between biological male and female brains by seeing patterns in structure and complexity invisible to the human eye.
These findings were validated using three different AI models designed to identify biological sex using their relative strengths by zeroing in on small areas of white matter or relationships in larger brain regions. can be analyzed.
“Our findings provide a clearer picture of how a living, human brain is structured, which in turn may provide new insights into how many psychiatric and neurological disorders develop,” said the study's senior author and neuroradiologist. and why they may present differently in men and women”. Yvonne Louie, MD.
Lui, a professor and vice chair for research in the Department of Radiology at NYU Grossman School of Medicine, notes that previous studies of brain microstructure have mostly relied on animal models and human tissue samples.
In addition, the validity of some of these past results has been called into question for relying on statistical analyzes of “hand-drawn” regions of interest, meaning that researchers had to know the shape, size, and location of the regions. I need to make many subjective decisions. They choose. Lowy says such choices can potentially skew results.
The results of the new study, published online May 14 in the journal Scientific reportsthe authors say, avoided this problem by using machine learning to analyze entire groups of images without asking the computer to inspect a specific location, which helped overcome human biases.
For the research, the team began by feeding existing data examples of brain scans of healthy men and women to AI programs and telling the machine programs the biological gender of each brain scan.
Because these models were designed to use complex statistical and mathematical methods to get “smarter” over time, they eventually “learned” to distinguish biological sex on their own. . Importantly, Louie says, the programs were prevented from using the overall size and shape of the brain.
According to the results, all models correctly identified the gender of a subject's scan between 92% and 98% of the time. Several characteristics in particular helped the machines determine them, including how easily and in what direction water can move through brain tissue.
“These findings highlight the importance of diversity when studying human brain diseases,” said study co-lead author Jinbo Chen, MS, a doctoral candidate at the NYU Tandon School of Engineering.
“If, as has historically been the case, men are used as a normative model for various disorders, researchers may be missing out on critical insights,” said study co-lead author Vara Lakshmi Bayanagri. MS, who is a graduate research assistant at the NYU Tandon School, added. of engineering.
Bianagri cautions that while AI tools can report differences in the organization of brain cells, they cannot reveal which traits are more likely in which gender. She adds that the study classified sex based on genetic information and only included MRIs of sex-matched men and women.
According to the authors, the team next plans to explore the development of sex-related brain structure differences over time to better understand the environmental, hormonal and social factors that may play a role in these changes. .
Funding: Funding for the study was provided by National Institutes of Health Grants R01NS119767, R01NS131458, and P41EB017183, as well as United States Department of Defense Grant W81XWH2010699.
In addition to Lui, Chen, and Bayanagari, others involved in the study were NYU Langone Health and NYU researchers Sohae Chung, PhD, and Yao Wang, PhD.
About this AI and neuroscience research news
the author: Shira Pollen
Source: NYU Langone
contact: Shira Polan – NYU Langone
Image: This image is credited to Neuroscience News.
Original Research: Open access.
“Deep Learning with Diffusion MRI as an In Vivo Microscope Reveals Sex-Related Differences in Human White Matter Microstructure” by Yvonne Lui et al. Scientific reports
Abstract
Deep learning with in vivo microscopy-like diffusion MRI reveals sex-related differences in human white matter microstructure.
Biological sex is an important variable in neuroscience studies where sex differences in cognitive function and neuropsychiatric disorders have been documented.
Although gross statistical differences have previously been documented in macroscopic brain structures such as cortical thickness or region size, less is understood about sex-related cellular-level microstructural differences that provide insight into brain health and disease. can provide
Studying these microstructural differences between men and women paves the way for understanding mental disorders and diseases that manifest differently in different sexes.
Diffusion MRI is an important in vivo, non-invasive procedure that provides a window into the microstructure of brain tissue.
Our study develops multiple end-to-end classification models that accurately predict a subject's gender using volume diffusion MRI data and uses these models to identify white matter regions. which differ most between men and women. 471 male and 560 female healthy subjects (age range, 22–37 years) from the Human Connectome Project were included.
Fractional anisotropy, mean variance and mean kurtosis are used to capture the microstructure characteristics of brain tissue.
Diffusion parametric maps are registered to a standard template to reduce bias that may arise from macroscopic anatomical differences such as brain size and contour.
This study uses three major model architectures: 2D convolutional neural networks, 3D convolutional neural networks and Vision Transformer (with self-supervision).
Our results show that all 3 models achieve high sex-classification performance (test AUC 0.92–0.98) in all diffusion metrics indicating definite differences in white matter tissue microstructure between males and females.
We further use complementary model architectures to inform the pattern of detected microstructural differences and the influence of short-range versus long-range interactions.
Occlusion analysis with the Wilcoxon signed-rank test is used to determine which white matter regions contribute most to sex classification.
The results indicate that sex-related differences are reflected in both local properties as well as long-range interactions of global properties/tissue microstructure.
Our highly consistent findings across models provide new insights supporting differences between cellular-level tissue organization of male and female brains, particularly in the central white matter.