The exact function of the anterior cingulate cortex (ACC) is one of the largest riddles in cognitive neuroscience and a major challenge in mental health research. ACC dysfunction contributes to a broad spectrum of neurological and psychiatric disorders, such as depression, ADHD, Parkinson's disease, OCD and many others, but nobody knows what it actually does. Recently a new theory has been developed about ACC function; the HRL-ACC (Hierarchical Reinforcement Learning Theory of ACC). This theory proposes that the ACC selects and motivates high-level tasks based on the principles of hierarchical reinforcement learning. The ACC associates values with tasks, selects the correct tasks and applies control over other neural networks (such as the dorsolateral prefrontal cortex and basal ganglia), which execute the tasks. The values of these tasks are attributed based on "reward prediction error signals", which are sent from the midbrain dopamine system to the ACC. These signals can be recorded using scalp-EEG as an "event-related brain potential" (ERP), which is called the "reward positivity". Until this day, the exact origin of the reward positivity is not yet known. Studies have delivered strong indirect evidence that the reward positivity is generated in the ACC. However, there is an important lack of direct evidence to support this hypothesis. The goal of this study is to provide direct evidence that the reward positivity is generated in the ACC by letting a group of patients with refractory epilepsy perform the virtual T-maze task (which is known to elicit reward positivity) and simultaneously recording intracranial video-EEG.

This is a monocentric prospective interventional clinical trial that will include patients with refractory epilepsy and for whom a hospitalization at the Centre for Neurophysiological Monitoring (CNM) at the University Hospital Ghent is planned for invasive video-EEG recording. This hospitalization is part of a pre-surgical work-up to determine if the patient is eligible for resective surgery. It is important to note that the decision to plan the patient for invasive video-EEG monitoring is made by neurologists and neurosurgeons specialized in refractory epilepsy and is in no way connected to this study.

Every patient that will be admitted to the CNM for invasive video-EEG monitoring has a consultation with Prof. Dr. Alfred Meurs about one month before hospitalization. During this consultation, Prof. Meurs will ask if the patient is willing to participate in this study. If the patient agrees and has given informed consent, the first session of the study will take place directly after the consultation. However, if the patient is not available at that time, a new appointment will be made for the first session. The first session has to take place before admission to the CNM.

The study consists of two experimental sessions. The first session takes place after giving informed consent and before admission to the hospital for invasive video-EEG monitoring. Therefore, this session is completely study-specific and may require and extra visit to the hospital if the patient is not available right after the consultation with Prof. Meurs.

During this session, the patient will perform the virtual T-maze task for 20 minutes while scalp-EEG will be recorded simultaneously. During this task, patients will see a T-maze on a computer screen and have to choose at every junction whether they go left or right. Depending on the direction they choose, they get rewarded or not. This feedback will elicit the reward positivity.

The rationale for this first session is that it is important to know whether or not the patient produces a normal reward positivity measured by scalp EEG, before the intracranial electrodes are placed.

The second session takes place during the admission at the CNM, after the intracranial electrodes are placed. The invasive video-EEG monitoring is part of a pre-surgical work-up and is planned independently from the study. During this session, the patient will once again have to perform the virtual T-maze task while scalp-EEG and intracranial-EEG will simultaneously be recorded.

After the two sessions, the data will be analysed using ERP analysis and the amplitude of the reward positivity will be calculated for every electrode. It is expected that the amplitude of the reward positivity ERP-component will be largest for patients with electrodes in the frontal lobe, located closely to the ACC. This will deliver direct evidence that the reward positivity is in fact generated in the ACC.

Inclusion Criteria:

• Age 18 years or older
• Patients with refractory epilepsy
• Planned hospitalization at CNM for invasive video-EEG recording with subdural, epidural and/or depth electrodes
• Patients have to be able to give informed consent themselves
• Imaging after placement of electrodes (CT and/or MRI) has to be available

Exclusion Criteria:

• Patients known with dementia or severe cognitive and/or psychiatric disorders that makes it impossible for patients to participate in this study
• Active alcohol and/or drug abuse