Secondary Logo

Journal Logo

Abstracts and Programme: EUROANAESTHESIA 2011: The European Anaesthesiology Congress: Monitoring: Equipment and Computers

Application of tsallis entropy to quantify burst suppression in rabbits under propofol anesthesia


Silva, A.; Campos, S.; Monteiro, J.; de Pinho, Guedes P.; Antunes, L.

Author Information
European Journal of Anaesthesiology: June 2011 - Volume 28 - Issue - p 30
  • Free

Background and Goal of Study: Burst suppression (BS) activity of the electroencephalogram (EEG) reflects deep anesthesia and has been used to help intra operative anesthetic depth monitoring. It appears as a transitional pattern between isoelectricity and high-frequency EEG and its quantification is normally based in the percentage of isoelectricity in EEG epochs1.

Tsallis entropy (TE) has been recently applied to quantify the presence of BS in the EEG of rats after cardiac arrest2, but there is no data on this analysis during anesthesia. The use of a non-linear analysis based method (TE) could be a more practical alternative to the traditional BS ratio, as the first one does not need the set of an amplitude limit to consider the signal as isoelectric that may vary with amplitude and recording conditions. In the present study, the potential to use TE to quantify BS was accessed using EEG recorded in rabbits under propofol anesthesia.

Materials and Methods: Six rabbits were anesthetized with three propofol infusion rates: 70, 100 and 130 mgkg-1h-1, each maintained for 30 minutes, in a random order for each animal. Data recording was performed in the awake animals, 20, 25 and 30 minutes after beginning each infusion rate, in the recovered animals and consisted in:

1) Electroencephalogram recordings;

2) Evaluation of depth of anesthesia according to a clinical scale;

3) Arterial blood samples for plasma propofol determination. Tsallis entropy and the classic burst suppression ratio were analyzed. Performance of the indexes was compared by prediction probability (Pk) and pharmacodynamic analysis. The Wilcoxon Matched pairs test was used to compare the Pk values of TE and BS.

Results and Discussion: A higher mean Pk value was obtained for BS, although not statistically significant (BS 0.78±0.9; TE 0.59±0.7). Pharmacodynamic modeling revealed a better coefficient of determination for BS with the Emax stimulation curve (R2=0.90) while TE had a lower R2 which the Emax inhibitory curve (0.34).

Conclusions: The classic linear BS quantification method derived from the EEG shows better capacity to detect anesthetic depth and better correlation with drug dose in rabbits.


1. Rampil, I. J. Anesthesiology 1998, 89, (4), 980-1002.
2. Dandan, Z.; Jia, X.; Ding, H.; Ye, D.; Thakor, N. V. IEEE Trans Biomed Eng 2010, 57, (4), 867-74.
© 2011 European Society of Anaesthesiology