Computational Physiology Laboratory

Center for Biomedical Engineering

University of Kentucky

  Sleep and EEG Research
  By analyzing data from polysomnograms (overnight recordings of sleep-related signals such as the EEG) we hope to develop ways to characterize EEG signals that will discriminate patients from normal subjects. We also want to develop analyses that identify changes in EEG signals that occur before an arousal (and may provide insights about the cause of the arousal).

Study objectives: (1) to determine if the degree of irregularity (measured by Sample Entropy, SpEn) of an EEG signal varies systematically with sleep state (Wake, non-REM state 2, non-REM state 3/4, Rapid Eye Movement (REM) state); (2) to determine if an increase in SpEn of the EEG occurs prior to an arousal; (3) to determine whether SpEn of the EEG is different from normal in patients with sleep apnea.

   

  Below: 10-min segment of a polysomnogram, showing the following signals: 2 EEGs (2nd and 4th traces from top), blood oxygen level (top trace) and respiration (bottom 2 traces). Note periodic breathing on the right side of the record.
 
Below: The nature of an EEG signal changes with the sleep state of the subject, as seen in these 30-sec tracings. Sample Entropy (SpEn) reflects the "irregularity" of the EEG signal.
Return to top of page View Document