Track genomics, neuroscience and molecular biology.
Search papers, labs, and topics across Lattice.
This paper addresses the underperformance of Brain Foundation Models (BFMs) compared to kernel ridge regression (KRR) on functional connectivity (FC) for predicting individual phenotypes from fMRI data. The authors demonstrate that KRR's success stems from its implicit eigenvalue weighting of FC, which is miscalibrated for phenotype prediction, and introduce a spectral filter to recalibrate these eigenvalues. By pretraining a compact encoder model using this recalibrated FC as a target, they achieve performance comparable to larger BFMs with significantly fewer parameters, outperforming FC on short scans and smaller cohorts.
A simple spectral filter on functional connectivity eigenvalues unlocks a new pretraining objective that yields a highly parameter-efficient fMRI encoder outperforming established brain foundation models.
Self-supervised pretraining reshaped prediction in language and vision, and brain foundation models (BFMs) inherited its promise. Representations learned from large unlabelled corpora should capture individual functional dynamics and generalise across cohorts. However, kernel ridge regression (KRR) fitted on functional connectivity (FC) matrices still predicts individual phenotypes more accurately than any BFM we tested. In this paper, we show that KRR is weighted by the eigenvalues of the FC which are miscalibrated for phenotype prediction. We apply an efficient spectral filter to recalibrate the eigenvalues of each subject's FC matrix, enabling the model to exploit more inter-individual variance. Across the 5 datasets, 11 parcellations and 6 prediction targets we tested, we match or exceed the KRR baseline. Based on this finding, we then pretrain a small encoder model on about 4,000 hours of fMRI from 162 open datasets, whereby we align the pairwise similarities between the embeddings of recording snippets with those between the recalibrated connectomes. Our model performs on par with the best of the 6 published BFMs we tested while having an order of magnitude fewer parameters. Our encoder performs better than FC on short scans and in smaller cohorts, especially in fingerprinting. We release the pretrained model weights, the code and the pretraining data, preprocessed and parcellated.