EEG autoregressive model parameters can be used to classify the different levels of anesthesia in the steady state. Twenty-four 4-second records were used to train an 8-parameter EEG classifier to recognize each of 4 MAC levels (0, 1, 1.5, 2). Selecting 35 additional records at random provided data to test the classifier. On the three separate channels of EEG, the test classification indicated an average probability of .85 for the correct MAC level.
|Original language||English (US)|
|Issue number||3 SUPPL|
|Publication status||Published - 1979|
ASJC Scopus subject areas
- Anesthesiology and Pain Medicine