Abstracts

  Epilepsy affects 1% of the world's population, with 30% of these patients suffering from drug-resistant epilepsy. To reduce seizures, it is crucial to define the seizure onset zone for surgical removal. However, preserving cognitive functions in neighboring regions remains a challenge. Electrical stimulation mapping has been used for decades to protect language regions in epilepsy surgery, but the technique typically continues to rely on two-dimensional maps for planning. With the increasing use of stereoelectroencephalography (SEEG), it is now possible to investigate cognitive functions deep in the brain. We aim to develop a three-dimensional (3D) model of functional organization in the human brain. Such maps would significantly improve pre-surgical planning, help predict cognitive outcomes, and provide a detailed guide for assessing impaired functions in patients following stimulation.

Data were collected from 23 epilepsy patients who underwent pre-surgical evaluation. This included CT and MRI scans, detailed records of SEEG implantation and stimulation sessions. CT and MRI scans were co-registered and SEEG coordinates were exported onto the MNI brain template to establish a common reference frame. We generated clusters by grouping SEEG contacts that were stimulated within a 10mm³ region across different patients. Then each cluster were assigned to its corresponding brain region using an atlas (AALS 3). This was then used to generate 3D model maps for 5 modalities (including visual and auditory naming, affective, language and sensory processing).

We identified and analyzed clusters of brain regions implicated in cognitive disruptions following electrical stimulation. Language processing disruptions were observed in 8 clusters, in the middle and superior temporal gyri, Rolandic operculum, postcentral gyrus, and temporal pole, affecting 5 patients. Visual naming disruptions were associated with 12 clusters, involving regions such as the frontal inferior triangularis, postcentral and supramarginal gyrus, superior, middle and inferior temporal gyri, occipital mid gyrus, insula, hippocampus, and frontal inferior operculum, affecting 6 patients. Similarly, we analyzed auditory naming, affective and sensory processing, identifying 2, 3, and 6 clusters respectively, across 4 patients for each modality.

3D brain maps visualize cognitive functions and provide reference points for surgery, offering clinical specialists insights into critical brain functions and surgical risks. Further cognitive domains could be explored, especially around memory. As more patients are included, additional clusters are likely to be identified, leading to a more precise mapping of the brain regions involved. This will also enable to discern patterns, such as differences in brain regions associated with specific tasks, like visual naming of objects versus visual naming of famous persons. Additionally, multisite collaborations to expand the number of patient included would be beneficial, allowing for further standardization of this project.