Abstracts

Rationale: Vertex Sharp Transients (VSTs) are electronegative waves that are electroencephalographic (EEG) hallmarks of non-REM sleep. Beyond its EEG features, little is known about the regions and networks involved in VSTs occurrence. Recordings from cat brains have identified corresponding activity in the deep medial regions of frontal, central, and parieto-occipital cortex. A similar study with Rhesus monkeys suggests an anterior limbic origin. In humans, magnetoencephalography has shown inferior parietal lobe involvement. There are no simultaneous EEG/fMRI (SEM) studies localizing signal changes during VSTs occurrence. Our purpose is to identify brain regions with fMRI signal change correlated to spontaneous VSTs. Methods: We reviewed EEGs recorded with SEM from participants with epilepsy studied at the UCLA Seizure Disorder Center over the last 4 years according to an IRB approved protocol. A total of 81 EEGs from 24 participants were included. VSTs were identified and distinguished from similar vertex activity indicating K-complexes and sleep spindles. fMRI data analysis was performed using fMRI Software Library (Oxford, UK) by convolving the VSTs-occurrence time model with a double-gamma hemodynamic response function. Higher level analysis was performed by using a mixed effects approach. Results: Clear VSTs were present in 4 10-minute EEGs from 3 participants with each EEG including 2, 5, 6, and 10 VSTs. Including this total of 23 VSTs, fMRI analyses demonstrated signal increase bilaterally in the cingulate cortex, paracentral frontal and parietal lobes, insula, and thalamus. Included images are thresholded at Z≥2.3 and superimposed on a standard brain template. Conclusions: Our results corroborate the findings in earlier animal recordings that suggest medial brain regions involvement in VSTs. These results also are consistent with the fMRI results from an analysis of non-REM sleep without specific modelling of non-REM transients. However, the use of SEM has allowed us to visualize a broader anatomic region than earlier physiological experiments with modelling of specific activity in non-REM sleep. We observed regions corresponding to VSTs occurrence that included several components of the limbic system, which is the first observation of this widespread network's correspondence to VSTs. The specific inclusion of the thalamus and cingulate cortex indicates involvement of regions recognized to be components of systems for alertness and attention, which may suggest a possible mechanistic role for VSTs. However, the additional limbic components may suggest a broader explanation of VSTs, which may be as an epiphenomenon for a mechanism that modulates alertness or attention and possibly also emotion and memory for a stable sleep state.