Different measures of brain connectivity answer different questions. Understanding their distinctions helps us design and interpret brain stimulation experiments.
Structural connectivity: possible anatomical routes
Diffusion MRI measures the effects of water diffusion on the MR signal. Diffusion tensor imaging (DTI) characterises its dominant orientation within a voxel; tractography uses diffusion information to estimate possible white-matter pathways.
These estimates provide anatomical context for a network. They do not directly show the direction of information flow or establish whether stimulation will change activity in another region.
Functional connectivity: signals that vary together
Functional connectivity describes statistical relationships between signals from different brain regions. In resting-state fMRI, researchers analyse spontaneous BOLD fluctuations; task-based studies examine activity in a defined cognitive context.
Correlated signals can be useful for describing networks and informing target selection. Correlation alone does not identify a direct pathway or establish which region influences another.
Effective connectivity and dynamic causal modelling
Dynamic causal modelling (DCM) fits models of interacting neural populations and their relationship to measured signals. For fMRI, this includes a model of how neural activity gives rise to haemodynamic responses. Comparing candidate models can test hypotheses about directed interactions and their modulation by experimental conditions.
These conclusions depend on the model, the regions included and the quality of the data. Model-based inference and experimental perturbation offer complementary evidence.
TMS-fMRI: introduce an intervention
Concurrent TMS-fMRI adds a controlled perturbation at a cortical target and measures the associated BOLD response across the brain. Appropriate comparison conditions help distinguish stimulation-related responses from sensory and other effects. A remote response is not, by itself, proof of a direct anatomical connection.
Chronometric TMS-fMRI further varies the timing of stimulation relative to a task. This allows us to investigate how network responses depend on cognitive state and intervention timing.
Observe functional connectivity, map responses to TMS, and explore task-relative timing.