Despite aggressive medical therapy, over one-third of the world’s 70 million epilepsy patients suffer from uncontrolled seizures. Surgery and implantable devices can control seizures in many patients, but these treatments are only effective when targeted accurately. Currently, targeting is manual due to the lack of rigorous methods to quantify epileptic networks, and when these targets are identified, there is no rigorous way to select the best surgical approach. Consequently, therapy varies dramatically across centers and patients. There is a critical need for standardized, quantitative methods to map epileptic networks and to target and optimize therapy. My long-term goal is to develop these quantitative methods, create a scalable infrastructure to implement them at scale, and rigorously validate and translate these methods into clinical practice. My overall objective is to integrate non-invasive structural imaging, which provides a comprehensive anatomical view of the brain, with invasive IEEG that aims to pinpoint seizure origin and spread. I will develop rigorous, quantitative methods to map epileptic networks that cause seizures to guide epilepsy surgery. My central hypothesis is that patient outcome after epilepsy surgery depends on what percentage of abnormal regions quantified on neuroimaging and IEEG are removed. With this central hypothesis, I will develop tools for clinical translation by (1) developing standardized quantitative methods that generalize across epilepsy centers, (2) developing and validating new methods to integrate structural imaging and IEEG, (3) implementing methods to run at scale on a large number of patients, representing the diversity of epilepsy, across centers. In Aim 1, I will develop scalable methods to automate aggregation and multimodal analysis of structural imaging, IEEG, and clinical data from multiple epilepsy centers. In Aim 2, I will develop normative methods that merge structural imaging and IEEG data, to identify abnormal epileptic networks by comparing individual patient’s data with the norm. Undertaking Aims 1 and 2 during the K99 phase will enhance my proficiency in cloud computing, scalable analysis, multicenter biostatistics, and clinical translation. In Aim 3, I intend to implement quantitative methods to run at scale to predict surgical outcomes in two specific populations: patients with temporal and extratemporal lobe epilepsy. Multiple conceptual and technical innovations are embedded in this proposal to overcome translational barriers that limit generalization, rigorous validation, and scalability. These include innovative tools to scale analysis, novel personalized localization methods, collaborative validation, and data sharing across 15 US epilepsy centers. This work is significant because it merges state-of-the-art engineering, neurology, and neurosurgery to make practical tools to improve and standardize patient care by quantitatively guiding epilepsy surgery.