Active Brain Regions During Sleep Using Simultaneous EEG-fMRI

Friday, 13 February 2015
Exhibit Hall (San Jose Convention Center)
Grace Y. Duan, Illinois Mathematics and Science Academy, Aurora, IL
With techniques allowing the simultaneous acquisition of electroencephalogram (EEG) and functional magnetic resonance imaging (fMRI) data, active brain regions during different stages of the human sleep cycle were determined. This allows a greater understanding of sleep and its role in processing information. Data were collected using a Siemens 3T TIM Trio magnet and Neuroscan Maglink 64 channel MRI compatible EEG. A standard EPI BOLD sequence was acquired for 45 minutes while the subject slept. Sequence parameters were TR=2s, TE=20ms, thirty two 1.7 mm in plane and 3 mm thick slices. The EEG data oscillations were used to determine the duration of each sleep stage. The respective MR images were grouped by sleep stage, motion corrected, and spatially smoothed. Specific regions of interest (ROIs) were extracted to create correlation maps, visualized using Mango software. Correlation matrices were created using MATLAB to display neuronal connectivity among all ROIs. Results suggest that brain activity is localized in the ROIs during sleep stage two. In rapid eye movement (REM) sleep, brain activity associated with the selected ROI is also present in most other brain regions. Hippocampal activity is most prevalent in all investigated stages, followed by thalamus, posterior cingulate, and amygdala activity. Brain activity in stage one and REM is similar, although REM activity is less concentrated to the ROI. The correlation matrices indicate that most brain regions show weak correlation with one another in sleep stages one and two, whereas in REM sleep, most brain regions are strongly associated with the others. Stage two is considered deeper sleep than stage one, showing diminished brain activity. REM sleep shows the most neuronal connectivity because it involves consolidating memories and dreaming, which requires much regional interaction. EEG-fMRI data can lead to an improved understanding of brain activity during sleep, which may help in diagnosing sleep disorders.