Over 15 million people worldwide suffer from localization-related drug-resistant epilepsy. These patients are candidates for targeted surgical therapies such as surgical resection, laser thermal ablation, and neurostimulation. While seizure localization is needed prior to surgical intervention, this process is challenging, invasive, and often inconclusive. In this work, I aim to exploit the power of multimodal high-resolution imaging and intracranial electroencephalography (iEEG) data to map seizure networks in drug-resistant epilepsy patients, with a focus on minimizing invasiveness. Given compelling evidence that epilepsy is a disease of distorted brain networks as opposed to well-defined focal lesions, I employ a graph-theoretical approach to map structural and functional brain networks and identify putative targets for removal. The first section focuses on mesial temporal lobe epilepsy (TLE), the most common type of localization-related epilepsy. Using high-resolution structural and functional 7T MRI, I demonstrate that noninvasive neuroimaging-based network properties within the medial temporal lobe can serve as useful biomarkers for TLE cases in which conventional imaging and volumetric analysis are insufficient. The second section expands to all forms of localization-related epilepsy. Using iEEG recordings, I provide a framework for the utility of interictal network synchrony in identifying candidate resection zones, with the goal of reducing the need for prolonged invasive implants. In the third section, I generate a pipeline for integrated analysis of iEEG and MRI networks, paving the way for future large-scale studies that can effectively harness synergy between different modalities. This multimodal approach has the potential to provide fundamental insights into the pathology of an epileptic brain, robustly identify areas of seizure onset and spread, and ultimately inform clinical decision making.