Abstracts

EEG-FMRI DEMONSTRATES ACTIVATION OF DIFFERENT THALAMIC NUCLEI IN IDIOPATHIC GENERALIZED EPILEPSY

Abstract number : 3.126
Submission category : 5. Human Imaging
Year : 2008
Submission ID : 8940
Source : www.aesnet.org
Presentation date : 12/5/2008 12:00:00 AM
Published date : Dec 4, 2008, 06:00 AM

Authors :
Jean Gotman, Louise Tyvaert, C. Hawco, S. Chassagnon, A. Sadikot, Pierre LeVan and F. Dubeau

Rationale: The thalamus is a complex structure made of several nuclei involved in different brain functions. It plays a major role in the generation of spike and wave discharges (SWD) in idiopathic generalized epilepsy (IGE) and usually with an increase BOLD response (activation) observed on EEG-fMRI. Until now, EEG-fMRI studies performed in IGE patients did not attempt to separate the BOLD responses within the thalamic nuclei. We studied the time course of the BOLD signal related to SWD in the centromedian parafascicular (CMPF) and anterior thalamic nuclei both involved in IGE patients with absences. Methods: Ten patients with IGE and absences were recorded using 3T EEG-fMRI during generalized SWD. Timing and duration of the SWD were integrated in the general linear model. For each patient, four regions of interest were defined corresponding to the anterior thalamic and CMPF nuclei of each thalamus (2x2x2 voxels of 5mm). Hemodynamic response functions (HRFs) were calculated for each region of interest using a Fourier basis set. The time to peak (positive and negative peaks) of HRF was analyzed in each case according to the signal-to-noise ratio. We compared these results within (left compared to right) and between (anterior thalamic compared to CMPF nucleus) each structure. Results: Four patients had a positive or a negative HRF peak in one of the regions of interest with a signal to noise ratio superior to 3. In these cases BOLD responses within each structure were symmetrical (p>0.3). We observed a delay of 1.25 ± 0.9 s (p=0.025) for the positive HRF peak and of 4.3 ± 1.5 s (p=0.001) for the negative peak between the CMPF and the anterior thalamic nucleus. In each case, positive and negative peaks occurred earlier in the CMPF. For patients (n=6) with BOLD responses and a signal to noise ratio of 1.5 to 3, we still observed symmetrical BOLD responses within each structure (p>0.7). A delay of 3.5 ± 3 (p=0.008) was again observed between the positive peak of the CMPF (earlier) and of the anterior thalamic nuclei. No delay was found for the negative peak. Conclusions: We found a delay in the BOLD time course between the CMPF and the anterior thalamic nuclei during SWD confirming that these nuclei are involved differently in SWD generation. It is believed that two different thalamic circuits are activated during generalized SWD: one includes the CMPF, brainstem, striatum and cortex, and the other involves the anterior thalamic nucleus, the mammillary bodies and the rest of the limbic system. Because the responses in CMPF nucleus occurred earlier, we can state with confidence that the CMPF circuit is not driven by the other one. The circuit including the CMPF may be involved in discharge initiation and propagation, while the anterior thalamic nucleus circuit may play a role in discharge maintenance. These findings also emphasize the importance to distinguish these nuclei in fMRI studies of other thalamic activities such as spindles. Supported by CIHR 38079
Neuroimaging