Print Close

CentileBrain Normative Models of Brain Morphometry Can Generalize to Diverse Ethnoracial Groups

Poster No:

1320

Submission Type:

Abstract Submission

Authors:

RUIYANG GE¹, Yuetong Yu¹, Guoyuan Yang², Jia-hong Gao³, Ryota Hashimoto⁴, Masaki Fukunaga⁵, Junya Matsumoto⁴, Kiyotaka Nemoto⁶, Neda Jahanshad⁷, Paul Thompson⁸, Sophia Frangou¹

Institutions:

¹University of British Columbia, Vancouver, British Columbia, ²Beijing University of Technology, Beijing, China, ³Peking University, Beijing, Beijing, ⁴National Center of Neurology and Psychiatry, Kodaira, Japan, ⁵National Institute for Physiological Sciences, Okazaki, Aichi, ⁶Department of Psychiatry, University of Tsukuba, Ibaraki, Japan, ⁷Imaging Genetics Center, Keck School of Medicine of University of Southern California, Los Angeles, California, ⁸Imaging Genetics Center, Keck School of Medicine of University of Southern California, Los Angeles, CA

First Author:

RUIYANG GE
University of British Columbia
Vancouver, British Columbia

Co-Author(s):

Yuetong Yu
University of British Columbia
Vancouver, British Columbia

Guoyuan Yang
Beijing University of Technology
Beijing, China

Jia-hong Gao
Peking University
Beijing, Beijing

Ryota Hashimoto
National Center of Neurology and Psychiatry
Kodaira, Japan

Masaki Fukunaga
National Institute for Physiological Sciences
Okazaki, Aichi

Junya Matsumoto
National Center of Neurology and Psychiatry
Kodaira, Japan

Kiyotaka Nemoto
Department of Psychiatry, University of Tsukuba
Ibaraki, Japan

Neda Jahanshad, PhD
Imaging Genetics Center, Keck School of Medicine of University of Southern California
Los Angeles, California

Paul Thompson, PhD
Imaging Genetics Center, Keck School of Medicine of University of Southern California
Los Angeles, CA

Sophia Frangou
University of British Columbia
Vancouver, British Columbia

Introduction:

Application of the normative modeling to brain morphometry data has the potential to inform about the significance of normative deviation in brain morphometry for health and disease [1,2]. However, the majority of data collected are from individuals of European ancestry or those who may self-identify as non-Hispanic Whites [3], and the generalization of the fit of established brain morphological norms to a broad range of ethnoracial groups remains largely unexamined.

Methods:

We have developed robust age- and sex-specific CentileBrain normative models (https://centilebrain.org/) [4] of regional measures of cortical thickness, cortical surface area, and subcortical volumes using structural brain scans from a pooled sample of 37,407 healthy individuals (53.33% female; aged 3 to 90 years) with diverse but mostly European background. In the present report, model parameters computed in the model development sample were tested for their generalizability to independent samples, each comprising healthy individuals either self-identifying as Black (n=284), South Asian (n=376), East Asian Chinese (n=1,136), and East Asian Japanese (n=970) or using genetic data to label ancestry as African (n=104), Admixed American (n=57), East Asian (n=415), and European (n=428). Mean-absolute-error (MAE) and root-mean-square-error (RMSE) served as the main measure of model performance.

Results:

Regardless of the definition of the ethnoracial groupings, the correlation coefficient between the MAE and RMSE values of regional morphometric measures obtained from the model development sample and the MAE and RMSE values of the corresponding morphometric measures obtained in each ethnoracial sample were all greater than 0.96 for all ethnoracial groups (Figure 1 and Figure 2).

·Figure 1

·Figure 2

Conclusions:

This study provides evidence that our pre-trained CentileBrain normative models for brain morphometric measures can be applied to samples of diverse ethnoracial backgrounds across the human lifespan.

Lifespan Development:

Normal Brain Development: Fetus to Adolescence ¹

Modeling and Analysis Methods:

Classification and Predictive Modeling ²

Novel Imaging Acquisition Methods:

Anatomical MRI

Keywords:

Aging

Computational Neuroscience

Development

Informatics

Morphometrics

STRUCTURAL MRI

^1|2Indicates the priority used for review

Provide references using author date format

1. Marquand, A., et al. (2016). Understanding Heterogeneity in Clinical Cohorts Using Normative Models: Beyond Case-Control Studies. Biological Psychiatry, 80, 552-561.
2. Haas, S., et al. (2023). Normative modeling of brain morphometry in Clinical High-Risk for Psychosis. JAMA Psychiatry, in press.
3. Palk, A., et al. (2020). Ethical issues in global neuroimaging genetics collaborations. NeuroImage, 221, 117208.
4. Ge, R., et al. (2023). Normative modeling of brain morphometry across the lifespan using CentileBrain: Algorithm benchmarking and model optimization. Lancet Digital Health (accepted).