Computerized EEG frequency analysis

Abstract

We performed computerized EEG frequency analysis (C-EEGFA) in 69 controls and 20 patients with focal brain lesions and focally abnormal conventional EEGs. Individual channel EEG frequency analysis variables that were helpful in differentiating the 2 groups were absolute delta and theta band power, relative delta, theta, and alpha band powers, and median-power frequency. High-frequency beta band power (20 to 32 Hz) was not useful. Changes in EEG with age were seen only after age 50 and generally consisted of an increase in anterior alpha power, with no significant increase in slowing. Correlations of C-EEGFA variables with posterior alpha power were more significant than correlations with age. Calculating normative C-EEGFA data for 5 subsets of controls, each with a different amount of posterior alpha power, increased the sensitivity of the EEG frequency analysis test without altering the specificity. Even with this correction 2 of 20 patients with focal lesions and focally abnormal conventional EEGs had normal C-EEGFA studies. If these obvious focal lesions produced normal results, more subtle diseases might not be detected. A significant clinical utility of C-EEGFA remains to be proven.