Sensory modality shapes post-learning functional connectivity.

Learning new categories through different sensory modalities induces distinct forms of neural plasticity that persist in resting-state brain networks. Here, we investigate how tactile versus visual category learning affects resting-state functional connectivity and network topography in 84 participants (41 visual and 43 tactile learners). Using correlation-based connectivity matrices and graph-theoretical analyses, we found that tactile learners exhibited higher network assortativity than visual learners. Local graph features analysis revealed cerebellar involvement in both modalities, while also indicating regional specialization for processing distinct sensory inputs. Significant between-group differences also emerged in the frontal regions, putamen, supramarginal gyrus, and parts of the temporal lobe. Finally, using the Fisher's score method and random forest classifier, we achieved 78.57% accuracy in distinguishing learning groups based on resting-state connectivity patterns. These findings demonstrate that sensory-specific learning creates lasting, modality-dependent imprints on brain network architecture, offering valuable insights into how the brain processes information across different sensory modalities.