Mapping Directional Connectivity in Resting-State fMRI Using Dynamic Causal Modeling in Matched Healthy Controls for Hypothalamic Hamartoma Patients

Hypothalamic hamartoma (HH)–related epilepsy is increasingly understood as a disorder of distributed neural networks rather than a focal lesion. While resting-state fMRI (rs-fMRI) has enabled mapping of functional connectivity within these networks, traditional approaches remain correlational and do not capture directionality. Dynamic Causal Modeling (DCM) addresses this limitation by estimating effective connectivity, allowing inference of directed interactions between brain regions. In this study, rs-fMRI data from 58 healthy controls were analyzed to establish normative effective connectivity within the hypothalamic–hippocampal circuit. Regions of interest included the hypothalamus and bilateral hippocampal heads, with connectivity parameters estimated using cross-spectral DCM. Results demonstrated consistent intrinsic self-inhibition across regions and modest, bidirectional extrinsic connectivity with continuous interindividual variability. Notably, hippocampal regions exerted a more stable influence on hypothalamic activity than the reverse, though overall directional asymmetry remained minimal. These findings define a normative baseline of directional connectivity, providing a critical reference for identifying pathological network alterations in HH-related epilepsy.