Abstracts

Rationale: Comorbidities between psychiatric disorders and patients with intractable epilepsy, particularly anxiety, show the intersection between seizure disorders and psychological state. Understanding the mechanism underlying this intersection could help alleviate symptoms in both epilepsy and psychiatric disorders. Intracranial recordings provide a unique opportunity to understand how behavioral correlates of anxiety and other disorders manifest in neural data. We can examine the neural correlates of anxiety and cognitive flexibility at the level of neural networks in conjunction with epilepsy. Our study is also unique in that we can analyze these neural and task data with self-report questionnaires and neuropsychological evaluations within the epilepsy group. This gives us better insight into the breadth of the psychiatric disorder for each participant and permits us to categorize neural or behavioral stereotypes with more precision. Our hypothesis was that those with psychiatric disorders would have an impaired ability to regulate emotional conflict, especially those who had high self-report and psychological evaluation scores for anxiety. 

Methods: In this study, we ran an emotional conflict resolution task (ECR) and Multi-Source Interference Task (MSIT) with three groups: epileptics who were undergoing clinically invasive electrode placement to determine seizure onset (n= 24), psychiatric patients with diagnosed disorders (n= 15), and healthy controls (n= 36). Both ECR and MSIT measure conflict resolution through changes in reaction time and accuracy for sequential presentations of congruent and incongruent trials, though ECR contains an emotional component while MSIT contains no emotional component (as a control). To link psychiatric state to the task, all participants filled out self-report psychological questionnaires as well as underwent neuropsychiatric testing for their clinical work-up. We also measured coherence between electrodes and brain regions. The reaction times, accuracies, and stimulus conditions were analyzed statistically for each participant.

Results: We show here that participants with epilepsy are less accurate in the ECR task than healthy controls (p< 0.05, Wilcoxon Rank Sum), though the participants do not have this accuracy deficit during the non-emotional MSIT task. We also see that a lower accuracy on the ECR task correlates with a higher score on measures of anxiety. These groups also have different network dynamics.