Abstracts

Rationale: When interictal epileptic discharges (IEDs) arise from the epileptic focus, electro-magnetic activity and metabolic-vascular changes occur (Heers, 2014). The latters have been mainly investigated by EEG-fMRI which is not sensitive to fluctuations in oxy-Hemoglobin, allows only short duration scans and has limited temporal resolution. We performed personalized and prolonged simultaneous EEG-NIRS recordings for monitoring local cortical fluctuations of deoxy (HbR) and oxy-Hemoglobin (HbO) concentrations at the time of IEDs (Peng, 2014). Our main goals were: 1) to validate the feasibility of long EEG-NIRS scans, 2) to characterize the Hemodynamic Response (HR) to IEDs, 3) to localize HR origin on the cortical surface. Methods: Four patients with neocortical focal epilepsy underwent a 3T MRI, MEEG (simultaneous Magneto and Electro-encephalography) and/or EEG-fMRI, which provided the spatial localization of the target region. This information was used to perform EEG-NIRS recording (up to 5 hours) spatially restricted to the epileptic zone and to the homologous contralateral region. NIRS optodes were glued on the scalp using collodion (Yücel, 2014) according to an individualized NIRS montage optimizing the sensitivity to the target area (Machado, 2013). IEDs were marked on EEG and isolated events (minimum of 35s from each-other) were selected. HR was estimated by averaging the corresponding HbR and HbO time-courses. Tomographic reconstructions along the cortical surface were computed using a Restricted Maximum Likelihood method with a minimum norm prior. Results: All patients tolerated well the procedure. Three patients showed a bilateral HbO increase, one a bilateral HbO decrease. All had a bilateral HbR decrease. For IEDs duration ranging 70ms to 2s, the overall HR peaked at about 15s (range 10-20s), which is longer than physiological HR. HbO increases paired to HbR decreases (Patients#1,2,4) are in agreement with an expected positive BOLD response (Fig.1). HbO and HbR decreases (Patient#3) suggest more surprisingly a reduction of regional cerebral blood volume (Fig.2). HbR and HbO 3D reconstructions confirm a large HR within the epileptic focus and within the contralateral homologous region. Conclusions: Individualized and prolonged EEG-NIRS recordings are feasible and allow characterizing and reconstructing HR to IEDs along the cortical surface. For the first time, we show that IEDs produce slow, bilateral and very long-lasting HbO and HbR modulations. The very long HR delay (all patients) and the decrease in cerebral blood volume (patient #3) are probably related to an impaired neurovascular coupling with possibly strong implications for HR modelling. References:Heers M et al., Hum Brain Mapp. 2014 doi:10.1002/hbm.22482. Machado A et al., J Biomed Opt. 2014;19(2):026010. Peng K et al., Epilepsy Res. 2014;108(3):491-505. Yücel MA et al, Neuroimage. 2014;85(1):192-201. Funding: 236625 NSERC Discovery Accelerator Supplement and 232798 Canadian Foundation for Innovation, operating fund 23739, CIHR MOP 201309