Abstracts

Rationale: Spindles and K-complexes are spontaneous EEG markers of non-rapid eye movement (REM) sleep that first appear in stage 2.They are believed to arise from thalamo-cortical networks involving thalamic reticular nucleus and they are likely related to sleep preservation.K-complexes may be elicited by environmental stimuli and spindles may inhibit perception of environmental stimuli.However,the regions generating the discharges and the functional connections of the generators to other regions are not fully known. Simultaneous EEG and fMRI provides depiction of brain regions with functional change during EEG events, and we used this technique to localize fMRI signal changes corresponding to spontaneous spindles and K-complexes. Methods: We reviewed 98 EEGs from 29 participants recorded during fMRI according to an IRB-approved protocol to investigate the anatomy of EEG activity.Spindles and K-complexes were identified and fMRI data analysis was performed using fMRI Software Library (FSL, Oxford, UK) by convolving the transients occurrence time with a double-gamma hemodynamic response function. Higher level group analysis was performed with a mixed effects approach. Images were thresholded at Z value of 2.3. Results: A total of 106 spindles and 68 K-complexes were identified within the EEGs of 7 participants who had spontaneous, deep sleep. Among the participants, 5 had epilepsy and 2 were controls without neurologic disease. The spindles appeared during 16 EEGs and the K-complexes during 18 EEGs.Image analysis of the spindles identified increased signal in bilateral thalamus, superior pre- and post-central cortex, posterior cingulate, pre-cuneus, and right superior temporal. Analysis of the K-complexes identified increased signal in bilateral thalamus, superior temporal, pre- and post-central cortex, and medial regions of the occipital, parietal and posterior frontal lobes. No regions of decreased signal occurred. Conclusions: The thalamic signal for both spindles and K-complexes is consistent with existing understanding of the role of each discharge in sleep preservation through gating sensory awareness. However, the thalamic finding for the K-complexes may still be related to the spindle associated with each K-complex because the fMRI temporal resolution does not allow differentiation of events occuring in the same TR. Each discharge also corresponds to signal subjacent to the EEG event, which most likely indicates a common neocortical generator. K-complexes include a broader extent of signal, encompassing primary sensory cortex for vision, touch, and hearing, and this may indicate an additional gating effect specific for the distinct sensory cortices. Asymmetric temporal signal corresponding to the spindles is unexpected and may be due to the contralateral temporal signal not reaching significance. Signal in this region had a Z value of 2.0. Additional investigation is needed to address this. Identification of the anatomic regions involved in spindle and K-complex generation may advance the understanding of thalamic networks and the electrophysiology of awareness during sleep.