FedFound: a federated foundation model for lifespan brain morphological connectome analysis
- 1Department of Radiology and Biomedical Research Imaging Center, University of North Carolina at Chapel Hill
Abstract
The brain morphological connectome derived from structural MRI reflects inter-regional morphological relationships, providing a powerful representation for characterizing individual variability and detecting abnormalities across the lifespan. However, these abnormal alterations are subtle and complex, posing significant challenges for accurate and generalizable diagnosis using machine learning. Here, we present FedFound, the first federated foundation model inspired by the structured educational and residency training pathway of radiologists, designed for robust and scalable analysis of lifespan brain morphological connectomes. Integrating heterogeneous neuroimaging datasets across sites and disorders (22,911 subjects aged 0 to 100 years), FedFound combines self-supervised pre-training and supervised federated disease-specific refinement, supporting multidisciplinary knowledge aggregation through distributed optimization. Across nine diagnostic tasks spanning neurodevelopmental, neuropsychiatric, and neurodegenerative disorders, FedFound demonstrates superior performance and interpretability, revealing both shared and disorder-specific morphological patterns across etiologies. FedFound provides a robust foundation for lifespan neuroimage-based diagnosis that complements clinical expertise, while establishing a scalable and generalizable paradigm for integrating heterogeneous neuroimaging data across institutions, populations, and diseases to advance medical foundation models.
Related
- Brain age prediction from structural and functional connectivity across the lifespan using connectome-based predictive modeling
- AREION: A Foundation Model for Atlas-Agnostic Representation Learning of Brain Connectome
- Connectome analysis reveals brain-wide visual processing in Drosophila
- Intrinsic connectivity signatures of anxiety sensitivity from connectome based predictive modeling
- Transcriptomic reconstruction of an interregional molecular connectome in schizophrenia
BibTeX
@article{han2026fedfound,
title = {{FedFound: a federated foundation model for lifespan brain morphological connectome analysis}},
author = {Kangfu Han and Dan Hu and Ya Wang and Zhengwang Wu and Sheng-Che Hung and Li Wang and Weili Lin and Gang Li},
journal = {npj Digital Medicine},
year = {2026},
month = jun,
doi = {10.1038/s41746-026-02925-7},
url = {https://doi.org/10.1038/s41746-026-02925-7},
note = {CC BY},
}
RIS
TY - JOUR
TI - FedFound: a federated foundation model for lifespan brain morphological connectome analysis
AU - Kangfu Han
AU - Dan Hu
AU - Ya Wang
AU - Zhengwang Wu
AU - Sheng-Che Hung
AU - Li Wang
AU - Weili Lin
AU - Gang Li
PY - 2026
DA - 2026/06/01
JO - npj Digital Medicine
DO - 10.1038/s41746-026-02925-7
UR - https://doi.org/10.1038/s41746-026-02925-7
AB - The brain morphological connectome derived from structural MRI reflects inter-regional morphological relationships, providing a powerful representation for characterizing individual variability and detecting abnormalities across the lifespan. However, these abnormal alterations are subtle and complex, posing significant challenges for accurate and generalizable diagnosis using machine learning. Here, we present FedFound, the first federated foundation model inspired by the structured educational and residency training pathway of radiologists, designed for robust and scalable analysis of lifespan brain morphological connectomes. Integrating heterogeneous neuroimaging datasets across sites and disorders (22,911 subjects aged 0 to 100 years), FedFound combines self-supervised pre-training and supervised federated disease-specific refinement, supporting multidisciplinary knowledge aggregation through distributed optimization. Across nine diagnostic tasks spanning neurodevelopmental, neuropsychiatric, and neurodegenerative disorders, FedFound demonstrates superior performance and interpretability, revealing both shared and disorder-specific morphological patterns across etiologies. FedFound provides a robust foundation for lifespan neuroimage-based diagnosis that complements clinical expertise, while establishing a scalable and generalizable paradigm for integrating heterogeneous neuroimaging data across institutions, populations, and diseases to advance medical foundation models.
SN - 2398-6352
ER -