Abstracts

Rationale: The correct identification of anatomic targets for responsive neurostimulation therapy (RNS) depth leads represents a crucial step for interfacing with a communicating epileptogenic circuit. Diffusion Tensor Imaging (DTI) has the potential to detect transient post-ictal associated changes in axonal water diffusion and define critical nodes in an epileptogenic circuit. Parametric Subtracted Post-Ictal DTI (pspiDTI) is introduced here as an imaging tool for detecting patient-specific ictal-related water diffusion abnormalities in white matter (WM) following stereotypic dyscognitive seizures without generalization secondarily for facilitating implantation of 1 to 2 4-contact RNS depth leads. This technique can identify inter- and post-ictal differences between fractional anisotropy (FA) and mean diffusivity (MD) measurements. Such differences can outline ictal-associated WM propagation pathways to identify critical nodes in a potentially extensive epileptogenic network for implantation up to two RNS depth leads. Methods: Six patients with medically intractable epilepsy, experiencing stereotypic dyscognitive seizures without generalization secondarily, who were candidates for RNS therapy were included in the study. Each patient underwent a post-ictal (up to 4hrs following seizure termination), and an inter-ictal (no electrocerebral seizures for at least 24hrs) DTI study. Imaging parameters consisted of 2mm-thick oblique slices in 60 non-collinear directions with a diffusion factor of 900 s/mm2 in a 3T scanner. Eddy current and motion corrections were applied. The resulting volumes were registered to a T1 MRI. Fractional anisotropy (FA) and mean diffusivity (MD) measures were then computed.Patient-specific voxel-wise t-tests were conducted on the resulting volumes to identify abnormal diffusivity regions. Statistically significant differences in FA and MD (p < 0.01) between the post-ictal and inter-ictal volumes were identified. Results: In 6/6 patients (100%), statistically significant (p < 0.01) regions of post-ictal FA decreases were detected. In 5/6 patients (83.33 %), each hyper-perfusion finding on ictal SPECT co-registered with MRI (SISCOM) spatially correlated with a significant FA decrease. Three of 6 (66.67 %) presented contralateral changes of FA without a previous transient hyper-perfusion finding on SISCOM in the contralateral hemisphere. In 5/6 patients (83.33 %), statistically significant post-ictal increases in MD were seen. Two patients presented a MD increase region that was anatomically near a transient hyper-perfusion region identified by SISCOM. The clinical relevance of abnormal diffusivity regions (FA and MD) complemented SISCOM and electroencephalographic data. This multimodal evaluation facilitated imaging seizure propagation pathways for incorporating into a presurgical planning system for defining targets for implanting a limited set of two 4-contact RNS depth leads. Conclusions: PspiDTI is an innovative post-processing neuroimaging technique that facilitates localizing abnormal WM regions of transient water diffusion involved in the activation of an epileptogenic network. When incorporated into a presurgical workflow, such regions, reflecting transient microstructural and physiological changes, can define critical nodes in a potentially extensive epileptogenic network with up to 3 epileptogenic sources for implanting a maximum of 2 RNS therapy depth leads. Funding: Mary Keane Fund, Foglia Family Foundation, CONACYT