From gradients to cognition: linking cortical manifolds to brain flexibility and disorder

Traditional neuroscience describes the cerebral cortex as a mosaic of discrete, functionally specialized regions. However, a complementary view has emerged, demonstrating that the brain is also organized along continuous gradients that capture large-scale transitions in connectivity, microstructure, and function. These gradients, derived using dimensionality reduction techniques on neuroimaging data, provide a low-dimensional manifold framework that unifies our understanding of how cortical architecture supports cognitive flexibility, learning, and clinical disorders. In this review, we integrate evidence from genetics, phylogeny, development, and multimodal neuroimaging to outline how macroscale gradients emerge from underlying biological constraints, become progressively decoupled from local microstructure in transmodal cortex, and dynamically reorganize during cognitive and clinical states. We further discuss how this framework provides new insights into individual differences, disease mechanisms, and recovery following brain injury. By bridging anatomy, function, and behavior, gradient-based approaches offer a powerful lens for mapping the architecture of human cognition and its disruption in disease.