Identification of epileptic networks with graph convolutional network incorporating oscillatory activities and evoked synaptic responses

Stereoelectroencephalography (SEEG) offers unique neural data from in-depth brain structures with fine temporal resolutions to better investigate the origin of epileptic brain activities. Although oscillatory patterns from different frequency bands and functional connectivity computed from the SEEG datasets are employed to study the epileptic zones, direct electrical stimulation-evoked electrophysiological recordings of synaptic responses, namely cortical-cortical evoked potentials (CCEPs), from the same SEEG electrodes are not explored for the localization of epileptic zones. Here we proposed a two-stream model with unsupervised learning and graph convolutional network tailored to the SEEG and CCEP datasets in individual patients to perform localization of epileptic zones. We compared our localization results with the clinically marked electrode sites determined for surgical resections. Our model had good classification capability when compared to other state-of-the-art methods. Furthermore, based on our prediction results we performed group-level brain-area mapping analysis for temporal, frontal and parietal epilepsy patients and found that epileptic and non-epileptic brain networks were distinct in patients with different types of focal epilepsy. Our unsupervised data-driven model provides personalized localization analysis for the epileptic zones. The epileptic and non-epileptic brain areas disclosed by the prediction model provide novel insights into the network-level pathological characteristics of epilepsy.

SEEG-based epileptic seizure network modeling and analysis for pre-surgery evaluation

Stereo-electroencephalography is a minimally invasive technique for patients with refractory epilepsy pursuing surgery to reduce or control seizures. Electrodes are implanted based on pre-surgery evaluations and can collect deep brain activities for surgery decisions. This paper presents a methodology to analyze stereo-electroencephalography and assist clinicians by recommending the optimal surgical option and target areas for focal epilepsy patients. A seizure network (graph) model is proposed to characterize the spatial distribution and temporal changes of ictal events. The network nodes and edges correspond to specific epileptogenic regions and propagation/impact pathways (weighted by directed transfer function), respectively. We then employ a K-means clustering strategy to group nodes into a few clusters, from which the target surgical areas can be identified. Ten patients with different types of focal seizures were thoroughly analyzed. Promising consistency between results of our method's recommendations, clinical decisions and surgery outcomes were observed.

Responsive Neurostimulation of the Mesial Temporal White Matter in Bilateral Temporal Lobe Epilepsy

BACKGROUND

Responsive neuromodulation (RNS) is a treatment option for patients with medically refractory bilateral mesial temporal lobe epilepsy (MTLE). A paucity of data exists on the feasibility and clinical outcome of hippocampal-sparing bilateral RNS depth lead placements within the parahippocampal white matter or temporal stem.

OBJECTIVE

To evaluate seizure reduction outcomes with at least a 1-yr follow-up in individuals with bilateral MTLE undergoing hippocampus-sparing implantation of RNS depth leads.

METHODS

A retrospective analysis of prospectively collected data was performed on patients at our institution with bilateral MTLE who were implanted with RNS depth leads along the longitudinal extent of bitemporal parahippocampal white matter or temporal stem. Baseline and postoperative seizure frequency, previous surgical interventions, and postimplantation electrocorticography and stimulation data were analyzed.

RESULTS

Ten patients were included in the study (7 male, 3 female). Overall seizure frequency declined by a median 44.25% at 3.13 yr (standard deviation 3.31) postimplantation. Four patients (40%) achieved 50% responder rate at latest follow-up. Two of four patients with focal onset bilateral tonic-clonic seizures became completely seizure-free. Forty percent of patients were previously implanted with a vagus nerve stimulator, and 20% underwent a prior temporal lobectomy. All depth lead placements were confirmed as radiographically located in the parahippocampal white matter or temporal stem without hippocampus violation. There were no cases of lead malposition.

CONCLUSION

Extrahippocampal or temporal stem white matter targeting during RNS surgery for bitemporal MTLE is feasible and allows for electrographic seizure detection. Larger controlled studies with longer follow-up are needed to validate these preliminary findings.

Neuromodulation of Epilepsy Networks

Neuromodulation, including first-generation open-loop devices and second-generation closed-loop devices, is a valuable but poorly understood therapeutic option for patients with drug-refractory epilepsy. The precise therapy a patient receives is contingent on the relationship between the patient's own unique neurophysiology and the custom programming of detection and stimulation parameters. Recent evidence demonstrates that therapeutic efficacy can be achieved through neuromodulation of seizure networks, rather than simple disruption of seizure evolution. Nevertheless, the improvement in outcomes achieved combined with its minimally invasive, nondestructive nature make closed-loop stimulation a promising therapy for additional indications, such as generalized and pediatric epilepsy.

Challenges of Epilepsy Surgery

Though frequently effective in the management of medically refractory seizures, epilepsy surgery presents numerous challenges. Selection of the appropriate candidate patients who are likely to benefit from surgery is critical to achieving seizure freedom and avoiding neurocognitive morbidity. Identifying the seizure focus and mapping epileptogenic networks involves an interdisciplinary team dedicated to formulating a safe and effective surgical plan. Various strategies can be employed either to eliminate the epileptic focus or to modulate network activity, including resection of the focus with open surgery or laser interstitial thermal therapy; modulation of epileptogenic firing patterns with responsive neurostimulation, deep brain stimulation, or vagus nerve stimulation; or non-invasive disconnection of epileptic circuits with focused ultrasound, which is also discussed in greater detail in the subsequent chapter in our series. We review several challenges of epilepsy surgery that must be thoughtfully addressed in order to ensure its success.

Networks in Frontal Lobe Epilepsy

Epilepsy affects about 1% of the general population. Frontal lobe epilepsy is the second most common focal epilepsy accounting for nearly 25% of medically refractory epilepsies. This paper reviews frontal lobe epilepsy from a perspective of a network disease that may help us to understand epilepsy from the microscale of genes, to local neuronal circuits, to the macrolevel of a whole-brain network. Surgical interventions, such as ablation and resection act by removing the active target nodes in the network, while responsive neurostimulation and vagus nerve stimulation act by modulating networks at the local neuronal circuit level and whole-brain level.