Back

Intracranial Stimulation in Epilepsy as a Novel Approach for Mapping Psychiatric Circuits

Sun, H.; Li, N.; Granados Martinez, A.; Chowdhury, F. A.; Diehl, B.; Akram, H.; Visser-Vandewalle, V.; Strange, B.; Barcia, J. A.; Polosan, M.; CHABARDES, S.; Giampiccolo, D.; Litvak, V.

2025-06-30 neurology
10.1101/2025.06.28.25330302 medRxiv
Show abstract

ObjectiveTo assess the feasibility of using stereo-electroencephalography (SEEG) depth electrodes, implanted for invasive recording of the epileptogenic zone in drug-resistant epilepsy, to identify and validate novel Deep Brain Stimulation (DBS) targets for psychiatric disorders. MethodsSEEG data from 86 patients with drug-resistant epilepsy implanted at the UK National Hospital for Neurology and Neurosurgery were analysed (6,372 electrode contacts). Normative structural and functional connectomes were combined with meta-analytic functional maps to estimate the probability of each contact engaging networks associated with psychiatric symptoms. A linear model, developed using published data from 80 patients with obsessive-compulsive disorder (OCD) treated with DBS, was used to validate the approach. ResultsNetworks involved in reward processing emerged as the most promising DBS targets, with a predicted required sample size of 7 patients (range 4-33). Bipolar disorder was the next most feasible (22 patients; range 10-44), followed by depression (32; range 16-49) and anxiety (38; range 17-59). For other disorders, estimated sample sizes exceeded 40, indicating limited feasibility for single-centre studies. ConclusionsThis study provides preliminary support for using SEEG to map brain circuits underlying psychiatric symptoms. The method leverages routinely implanted electrodes in epilepsy to explore candidate DBS targets, supporting a network-guided strategy for personalised neuromodulation in neuropsychiatry.

Matching journals

The top 7 journals account for 50% of the predicted probability mass.

50% of probability mass above

"Similar papers" are the closest papers from that journal in the model's embedding space. They show what the match is built on, but the ranking comes mostly from a classifier over the whole training set, not from these examples alone.