0127 Relationship Between Adenosine and Glutamate During Sleep Deprivation

Abstract

Introduction:The spatial resolution of fMRI allows for a detailed description of brain activity during sleep and presents a unique opportunity to gain a deeper understanding of the functions of sleep. However, the adverse sleep conditions of the MRI have made the continuous observation of an entire night's sleep opportunity difficult. We have assessed the feasibility of a method for promoting and measuring sleep in the MRI across an entire night. Methods: Volunteers for the two-night inpatient protocol were thoroughly screened with the use of psychometrically validated questionnaires to ensure they could sleep in the MRI environment. Prior to the inpatient visit, volunteers were habituated to an ~23:00-07:00 sleep schedule for 14 days, verified by actigraphy, and were given the opportunity to acclimate themselves to the MRI acoustic noise. Hearing safety was assessed by baseline and post-MRI audiology examinations. Volunteers slept on a clinical-grade memory foam mattress and were equipped with active noise cancellation headphones, in addition to standard hearing protection. EEG data acquired during scans were artifact-corrected in real-time to allow for sleep scoring. Other peripheral signals were measured such as variations in cardiac and respiratory cycles and eyelid position. Results: On the second night, seven of eleven volunteers were successfully recorded in the scanner for 317-505 minutes. Preliminary review of audiology data found no clinically relevant changes in puretone thresholds or early auditory evoked potentials. Real-time EEG scoring approximated that these subjects spent 45-240 minutes in SWS and 40-90 minutes in REM. EEG data were scored offline with optimized artifact correction algorithms in one volunteer so far. This volunteer exhibited 54 minutes of SWS, 39 minutes of REM, and 279 minutes of total sleep. Total recording time was 369 minutes, and 63 minutes were unscorable. Conclusion:These results demonstrate that all-night continuous fMRI scanning is a viable technique for measuring sleep. This opens many potential future directions for studying the neural circuits associated with sleep phenomena such as state cycling.