Summary
As aberrant network-level functional connectivity underlies a variety of neural disorders, the ability to induce targeted functional reorganization would be a profound development towards therapies for neural disorders. Brain stimulation has been shown to alter large-scale network-wide functional connectivity, but the mapping from stimulation to the modification is unclear. Here, we leverage advances in neural interfaces, interpretable machine learning, and graph theory to arrive at a model which accurately predicts stimulation-induced network-wide functional reorganization. The model jointly considers the stimulation protocol and the cortical network structure, departing from the standard approach which only considers the stimulation protocol. We validate our approach in the primary sensorimotor cortex of non-human primates using paired optogenetic stimulation through a large-scale optogenetic interface. We observe that the stimulation protocol only predicts a small portion of the induced functional connectivity changes while the network structure predicts much more, indicating that the underlying network is the primary mediator of the response to stimulation. We extract the relationships linking the stimulation and network characteristics to the functional connectivity changes and observe that the mappings diverge over frequency bands and successive stimulations. Finally, we uncover shared processes governing real-time and longer-term effects of stimulation. Our framework represents a paradigm shift for targeted neural stimulation and can be used to interrogate, improve, and develop stimulation-based interventions for neural disorders.
Competing Interest Statement
The authors have declared no competing interest.
