Autoregressive model order estimation criteria for monitoring awareness during anaesthesia

Nicoletta Nicolaou, Julius Georgiou

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

This paper investigates the use of autoregressive (AR) model order estimation criteria for monitoring awareness during anaesthesia. The Bayesian Information Criterion (BIC) and the Akaike Information Criterion (AIC) were applied to electroencephalogram (EEG) data from 29 patients, obtained during surgery, to estimate the optimum multivariate AR model order. Maintenance of anaesthesia was achieved with propofol, desflurane or sevoflurane. The optimum orders estimated from the BIC reliably decreased during anaesthetic-induced unconsciousness, as opposed to AIC estimates, and, thus, successfully tracked the loss of awareness. This likely reflects the decrease in the complexity of the brain activity during anaesthesia. In addition, AR order estimates sharply increased for diathermy-contaminated EEG segments. Thus, the BIC could provide a simple and reliable means of identifying awareness during surgery, as well as automatic exclusion of diathermy-contaminated EEG segments.

Original languageEnglish
Title of host publicationArtificial Intelligence Applications and Innovations - 9th IFIPWG 12.5 International Conference, AIAI 2013, Proceedings
Pages71-80
Number of pages10
DOIs
Publication statusPublished - 1 Dec 2013
Externally publishedYes
Event9th IFIP WG 12.5 International Conference on Artificial Intelligence Applications and Innovations, AIAI 2013 - Paphos, Cyprus
Duration: 30 Sep 20132 Oct 2013

Publication series

NameIFIP Advances in Information and Communication Technology
Volume412
ISSN (Print)1868-4238

Conference

Conference9th IFIP WG 12.5 International Conference on Artificial Intelligence Applications and Innovations, AIAI 2013
CountryCyprus
CityPaphos
Period30/09/132/10/13

Keywords

  • Anaesthesia
  • AR model order estimation
  • Awareness
  • EEG

Fingerprint Dive into the research topics of 'Autoregressive model order estimation criteria for monitoring awareness during anaesthesia'. Together they form a unique fingerprint.

Cite this