Project Summary/Abstract The most common drug resistant epilepsy involves the temporal lobe (TLE) and a quarter of patients continue to suffer from seizures after temporal lobectomy, with children designated as temporal plus epilepsy (TLE+) having a five-times increased risk of post-operative surgical failure (continued seizures). The long-term goal of our research program is to develop brain connectivity tools that optimize the use of targeted therapies, including surgery. The objective of the proposed study is to use visual analysis and connectomics of MEG virtual sensor waveforms to predict the presence of spikes on intracranial EEG (iEEG) and pre-surgically distinguish between patients with TLE and TLE+. The central hypothesis is that user-defined virtual sensor beamforming (UDvs- beamforming), using expert reader analysis and connectomics, are superior to ECD, current density source modeling, and conventional beamforming for non-invasively differentiating between TLE and TLE+. The rationale is that new, validated MEG methods would improve pre-surgical planning of iEEG and better identify patients at risk for worsened outcomes. The research proposed in this application is innovative because (1) while MEG connectivity measures have been studied in epilepsy, its use in pre-surgical evaluation for TLE/TLE+ would be novel, (2) our approach can characterize both local and widespread connectivity patterns without sampling bias, and (3) results can be collected on an individual patient basis which will facilitate immediate integration into existing clinical pipelines. This contribution is significant because it is the first step in a program of research that is expected to improve pre-surgical planning for patients with drug resistant epilepsy and improve understanding of treatment failure and outcomes in epilepsy.