Overall, this study will investigate the functional utility of stereotyped HFOs by capturing them with a new implantable system (Brain Interchange - BIC of CorTec), which can sample neural data at higher rates >=1kHz and deliver targeted electrical stimulation to achieve seizure control. In contrast to current closed-loop systems (RNS), which wait for the seizure to start before delivering stimulation, the BIC system will monitor the spatial topography and rate of stereotyped HFOs and deliver targeted stimulation to these HFO generating areas to prevent seizures from occurring. If the outcomes of our research in an acute setting become successful, the investigators will execute a clinical trial and run the developed methods with the implantable BIC system in a chronic ambulatory setting.

High-frequency oscillations (HFOs) of intracranial EEG (iEEG) have the potential to identify the surgical resection area/seizure onset zone (SOZ) in patients with drug-resistant epilepsy. However, multiple reports indicate that HFOs can be generated not only by epileptic cerebral tissue but also by non-epileptic sites often including eloquent regions such as motor, visual, and language cortices. In this project, the initial evidence of a recurrent waveform pattern is presented that may be sufficient to distinguish pathological HFOs from physiological ones. Specifically, the investigators show that the SOZ repeatedly generates sets of stereotypical HFOs with similar waveform morphology whereas the events recorded from out of SOZ were irregular. This morphological pattern served as a robust neurobiomarker to isolate SOZ from other brain areas in multiple patients consistently. While these promising preliminary results are in place, the functional utility of stereotyped HFOs in a closed-loop seizure control system remains unknown. As of today, not much is known whether the stereotyped HFOs generated by the SOZ can be detected with an implantable system. If this can be achieved, then HFOs can be strategically translated as a neurobiomarker into closed-loop seizure control applications. The investigators hypothesize that pathologic stereotyped HFOs can be captured with the implantable Brain Interchange (BIC) system of CorTec and spatial topography of these events can be utilized by the implantable system to deliver targeted electrical stimulation to achieve seizure control. Using an acute setup within the epilepsy monitoring unit (EMU), this project will investigate the feasibility of capturing stereotyped HFO events using the new BIC system and compare the detection results to those obtained with the commercially available amplifier (Aim-1). If the first phase (Aim-1) of the study becomes successful, later in the second phase (Aim-2), once again in the EMU, the investigators will deliver targeted electrical stimulation to those brain sites associated with stereotyped HFOs using the BIC.

During the entire study, the BIC system will not be implanted but used externally to assess the neural data through externalized electrodes and then deliver closed-loop stimulation. If the outcomes of the research in an acute setting become successful, the investigators will execute a clinical trial and run our methods with the implanted BIC system in a chronic ambulatory setting.

Inclusion Criteria:

• Patients with medically refractory epilepsy, who have been deemed appropriate candidates for intracranial EEG monitoring
• Adult men and women (18≤ age <70 years)
• Children (3≤ age <18 years)
• Includes women and minorities

Exclusion Criteria:

• Subjects will be excluded if their condition makes them unable to continue with recordings.

Intracranial EEG + ECoG recorded from patients with epilepsy will be shared. The electrical signs of neural activity will be collected in the EMU for 24hrs. One stream of data will be digitized at 1kHz with 16bit resolution with the BIC system and the other stream with least a 2 kHz sampling frequency to capture HFOs reaching up to 500Hz.

Additionally, the closed-loop stimulation data with stimulation parameters including onset, duration, frequency, pulse width, and amplitude information will be shared. All collected data will be de-identified prior to data exchange. Patient data will be provided in a coded format that protects patient identities but will contain diagnosis (signs/symptoms), interventions including technical observations, diagnostic tests/results, and patient outcomes.

Information about the device delivering therapy including device serial numbers, device model numbers, date of the event, and country/state of the event will be annotated with the data and therapy

Every 6 months

Data Archive for the BRAIN Initiative (DABI) - invasive human neurophysiology (including EEG, ECoG, LFP, single unit)