TIME-FREQUENCY SPECTRAL REPRESENTATION OF THE EEG AS AN AID IN THE DETECTION OF DEPTH OF ANESTHESIA

A time-frequency spectral representation (TFSR) has been used to study the nonstationary information in the EEG as an aid in determining the anesthetic depth. This paper uses a TFSR with an exponential weightin g function for the purpose. Raw EEG data were collected from 10 mongre l dogs at various levels of halothane anesthesia. Depth of anesthesia was tested by observing the response to tail clamping, which is consid ered a supramaximal stimulus in dogs. A positive response was graded a s awake (depth 0), and a negative response was graded as asleep (depth 1). The EEG obtained during a period of 30 sec tail clamp was process ed into TFSRs. It was observed that at depth 0, the spectrum becomes l ocalized in time and frequency. The percent age of energy in the delta (1-3.5 Hz) and theta (3.5-7.5 Hz) frequency bands increased. At depth 1, the spectrum remained unchanged throughout the period of tail clam p. The performance of the TFSR in detecting the patient's awareness wa s also compared with the power spectrum. It was concluded that under c ertain anesthetic conditions, the TFSR is superior to the power spectr um.