Quadrimodal localisation of epileptic focus using simultaneous EEG, (f)MRI and PET imaging

Purpose: In patients suffering from pharmaco-resistant focal epilepsy, resection of the epileptic focus can lead to seizure-freedom or significant improvement. The localisation of the epileptic focus relies on structural and functional brain imaging and multimodal concordance is associated with a better post-operative outcome. The existence of epileptogenic lesions detectable on structural MRI and the presence of focal hypometabolism on FDG-PET, especially after coregistration with MRI are widely accepted localising findings. In addition, other tools for localising epileptic activity, such as EEG-based electric source imaging (ESI) and simultaneous EEG and functional MRI (EEG-fMRI) are increasingly used. We here report the feasibility to record in a single session combined EEG-PET, MRI, EEG-fMRI and ESI using high-density EEG and a PET-MRI hybrid scanner.

Methods: Ten patients with pharmaco-resistant epilepsy benefited from this multimodal recording. EEG was recorded at 1 kHz using an MR-compatible high-density (256 electrodes) EEG system (Electrical Geodesics Inc., Eugene, Oregon). PET and MRI images were acquired using an hybrid PET/MR scanner (Philips Ingenuity TF PET/MR) consisting of a Time-of-Flight PET and a 3T MRI linked through a single rotating patient table, allowing for MRI and PET imaging in the same session.

Data acquisition:(1) Simultaneous EEG-fMRI was recorded during 20 minutes at rest with eyes closed. (2) During the 30 minutes of FDG-PET tracer uptake after injection, the patient was at rest under EEG monitoring in order to document potential subclinical epileptic seizures that could cause false normalization of the interictal metabolism. Fifteen minutes of PET scan were recorded. (3) A full MRI epilepsy protocol and attenuation correction (AtMR) were finally acquired in coregistration with the PET acquisition.

Analysis: EEG-fMRI: After the correction of EEG artefacts, spike-related haemodynamic changes and epileptic topography-related BOLD changes were obtained [1]. (2) PET: high-resolution Time-of-Flight reconstruction was performed using the coregistered MR-based attenuation correction (AtMR) [2]. (3) Visual detection of MR structural abnormalities. (4) EEG acquired at rest during PET-FDG tracer uptake was used for Electric Source Imaging (ESI) using an individual head model derived from the 3D-T1 and a linear distributed inverse solution [3].

Results: The whole multimodal recording could be performed in less than two hours of scanner time and was well tolerated by the patients. This single-session quadrimodal imaging provided reliable interictal clinical data with good patient comfort. In all patients, we obtained concordant results of at least two modalities that were also concordant with electro-clinical ictal semiology.

Discussion & Conclusion: This single-session could replace four separate exams usually performed during presurgical epilepsy evaluation, avoiding the irradiation associated with the usual PET-CT. Additional functional mapping of eloquent cortex (motor, language) could be added in individual cases with little duration increase. This multimodal imaging is crucial for localizing the epileptic focus and improving the surgical outcome. This new procedure improves the workflow and may decrease the cost of presurgical evaluations.

References: [1] Grouiller F. et al. (2011), Brain, 134(Pt 10):2867 – 86. [2] Garibotto, V. et al. (2013), Clin. Nucl. Med., 38(1):e13 – 8. [3] Brodbeck V. et al. (2011), Brain 134(Pt 10): 2887 – 97.