Benchmarking External Generalization of SPD Matrix Learning for Resting-State fMRI Connectome Prediction
2026-08-31 • Machine Learning
Machine Learning
AI summaryⓘ
The authors studied how well brain scan data used to predict age works when tested on data from different groups of people. They used six sets of resting brain scans and compared different methods that take into account the special math properties of brain connection patterns. They found that prediction works better when training and testing data come from the same or mixed groups, but accuracy drops when tested on entirely separate groups, especially if the age ranges differ. Their work also provides a shared way for other researchers to test their methods fairly on new data.
resting-state fMRIfunctional connectivityage predictionSPD matrixmanifold geometrycross-dataset validationRiemannian harmonizationGroupKFoldbenchmarkcohort heterogeneity
Authors
Ce Ju, Antoine Collas, Florent Bouchard, Bertrand Thirion
Abstract
Resting-state functional magnetic resonance imaging (rs-fMRI) functional connectivity (FC) matrices are widely used for individual-level prediction, but strong performance within one cohort may not generalize to a new cohort. We ask whether within-dataset performance remains when the test data come from an entirely held-out rs-fMRI dataset. Each scan is represented as a regularized symmetric positive definite (SPD) correlation connectome, which allows methods to use the geometry of the SPD manifold. We introduce a reproducible age-prediction benchmark across six rs-fMRI datasets: COBRE, ADNIDOD, Cam-CAN, ABIDE, OASIS-3, and ADNI. The benchmark compares a vectorized correlation baseline, Tangent-Space Ridge, SPDNet, and split-wise Riemannian harmonization under within-dataset GroupKFold, pooled GroupKFold, and leave-one-dataset-out (LODO) evaluation. Within-dataset and pooled GroupKFold results are substantially more favorable than LODO results. When an entire dataset is held out, prediction error increases, differences among methods narrow, and performance is strongly affected by age-range mismatch and cohort heterogeneity. The benchmark provides common inputs, model settings, data splits, and analysis scripts so that future SPD matrix learning methods can be evaluated under the same external-validation protocol.