We aim to determine the clinical utility of 'dynamic tractography': a novel method for visualizing electrical neural transfers that incorporates the underlying white matter tracts and supporting linguistic processing. We will also determine how well objective electrophysiology biomarkers will improve the prediction of language outcomes following epilepsy surgery. This project will ultimately optimize understanding of how the human brain develops its language network dynamics.

We will determine the utility of a novel brain mapping technique for epilepsy presurgical evaluation, referred to as 'six-dimensional (6D) dynamic tractography'. This innovative program animates the rapid neural propagations along MRI-defined, 3D white matter tracts that connect regions supporting cognitive functions. Specifically, it will use event-related high gamma activity to localize the regions supporting specific linguistic functions and compute the velocity and strength of neural propagations based on the latency and amplitude of early neural responses to single-pulse electrical stimulation. We expect that considering both the negative effect of damaged white matter tracts and the positive effect of seizure control will help optimize the model's performance in predicting postoperative language outcomes; this will be accomplished by incorporating the 6D dynamic tractography and objective epilepsy biomarkers, including spontaneous high-frequency oscillations (HFOs) coupled to slow-waves, into our predictive model. By also identifying and considering the physiological high gamma augmentation strictly time-locked to stimuli and behaviors, our innovative intracranial EEG analysis will better distinguish the randomly-occurring pathologic HFOs. Another significant advancement provided by our model is its independence of conventional electrical stimulation mapping, which can acutely elicit seizures and fail to satisfactorily localize language areas in certain patient subsets. Additionally, this project will use 6D dynamic tractography to provide an explicit neurobiological model of language network dynamics, allowing us to tease apart the specific pathways originating from temporal lobe cortices that support the lexical retrieval of auditory or visual domains. Our prior project indicated that the arcuate fasciculus fibers support the direct transfer of lexical representations of auditory sentences. We will now determine whether the lexical representations of visual objects are likewise transferred via the arcuate fasciculus or others, including the fusiform-parietal fasciculus. To accomplish these goals, this project will prospectively recruit a new cohort of 80 epilepsy patients - age range: 0.5 to 21 years - undergoing extraoperative intracranial EEG recording and subsequent resective surgery. Finally, we will determine if the human brain creates and strengthens language-related functional parcellations throughout development. It has been suggested that the adult brain efficiently activates the posterior superior-temporal gyrus (STG) only during sound onset to decode the boundary between sounds. In contrast, the anterior STG shows sustained activation during an auditory stimulus to encode the phonetic features. We will determine if such a functional parcellation is more evident in older individuals, whose brains are more developed. While providing hypotheses focusing on specific brain regions, we will perform all of the proposed analyses at the whole-brain level. We will make all data and codes publicly available to facilitate external validation and implementation.

Inclusion Criteria:

• Epilepsy surgery with extraoperative intracranial EEG recording with subdural or depth electrodes.

Exclusion Criteria:

• Patients with a malignant brain tumor, or progressive neurodegenerative or metabolic disorders who are not considered candidates for intracranial EEG recording.

Following data de-identification, we propose to share our datasets with the public via 'NeuroElectroMagnetic Data Archive and Tools Resource (NEMAR)' or 'iEEG.org', both of which are NIH-funded data-sharing platforms.

Once the primary findings have been published, we will make the data available.