Domain adapted brain network fusion captures variance related to pubertal brain development and mental health

• Published in: Nature communications Vol. 14; no. 1; pp. 6698 - 10
• Main Authors: Kraft, Dominik, Alnæs, Dag, Kaufmann, Tobias
• Format: Journal Article
• Language: English
• Published: London Nature Publishing Group UK 23.10.2023; Nature Publishing Group; Nature Portfolio
• Subjects: 59/57; 631/378/116; 631/378/2649; 631/477; Adaptation; Adolescent; Adolescents; Brain; Brain - diagnostic imaging; Brain research; Charts; Cognitive development; Datasets; Humanities and Social Sciences; Humans; Longitudinal Studies; Machine learning; Medical imaging; Mental disorders; Mental Health; multidisciplinary; Neuroimaging; Phenotypes; Psychopathology; Puberty; Science; Science (multidisciplinary); Similarity
• ISSN: 2041-1723; 2041-1723
• DOI: 10.1038/s41467-023-41839-w

Puberty demarks a period of profound brain dynamics that orchestrates changes to a multitude of neuroimaging-derived phenotypes. This complexity poses a dimensionality problem when attempting to chart an individual’s brain development over time. Here, we illustrate that shifts in subject similarity of brain imaging data relate to pubertal maturation in the longitudinal ABCD study. Given that puberty depicts a critical window for emerging mental health issues, we additionally show that our model is capable of capturing variance in the adolescent brain related to psychopathology in a population-based and a clinical cohort. These results suggest that low-dimensional reference spaces based on subject similarities render useful to chart variance in brain development in youths. Charting variance in brain development across adolescence remains a challenge. Here, the authors show that domain adapted brain network fusion based on subject similarities in brain structure is associated with puberty and mental health in two independent cohorts.