PROJECT SUMMARY The current clinical standard-of-care for neurosurgical interventions for drug-resistant epilepsy (DRE) involves the use of intracranial electrodes for localization of individual cortical sites responsible for seizures and for sensorimotor and language functions. This static localizationist approach has not kept pace with modern models of the human brain as spatiotemporally dynamic multistate system of interacting networks. Moreover, the current clinical approach has not achieved satisfactory rates of post-operative freedom from seizures or cognitive impairments. The overall goal of this renewal is to achieve these key conceptual (moving from localization-based to network-based approaches) and practical (to improve seizure and language outcomes after epilepsy surgery) objectives. During the previous funding period, we developed and validated methodologies to study language and epileptogenic networks in the brain including, high-gamma modulation (HGM) language mapping, spatiotemporal propagation of HGM, functional connectivity based on coherence modulations, and effective connectivity from cortico-cortical evoked potentials (CCEPs) elicited with single pulse electrical stimulations (SPES). We generated evidence that (1) task-related HGM can localize language regions with high specificity and improved safety compared to the current clinical standard of electrical stimulation mapping (ESM) and is a better predictor of postsurgical standardized language outcomes, (2) spatiotemporal propagation of HGM and coherence modulations can delineate neuronal circuits involved in language processing and their dynamical properties on different time-scales, (3) incorporating CCEPs to define current propagation along functional pathways can reconcile and improve sensitivity of HGM language mapping compared to ESM, (4) CCEPs can delineate patient-specific epileptogenic networks which can be harnessed for precise and personalized selection of surgical targets to improve