Abstract
Abstract: :
Purpose: To establish the contribution that changes in alertness make to response variability in a visual field type task using EEG frequency spectra as a measure of the state of alertness and the frequency-of-seeing curve (FOS) as a measure of response variability. Methods: In experiment 1, EEG frequency spectra for true positive and false negative responses (errors) were established in 3 subjects during a 5dB supra-threshold visual field type task with random insertion of blank trials (25%). 600 presentations were made in a single session lasting approximately 25 minutes. In experiment 2, subjects were presented with an FOS task while their occipital EEG was being monitored to identify periods of reduced attention. Responses were then excluded on the basis of the EEG data and FOS curves re-computed. Results: In experiment 1, two averaged EEG frequency spectra were computed for true positive and false negative responses. These showed marked differences at 2 frequencies. For false negative responses the amplitude at ~10Hz (alpha) reduced while that at ~5Hz increased. The exact frequency of the varying components was subject dependent. In experiment 2, the EEG frequency spectra during extended periods of non-response corresponded to those of false negative epochs in experiment 1 (increased 5:10Hz ratio). Exclusion of data based either upon periods of non-response or upon these EEG components altered the position and shape of the FOS curve. Conclusions: Periods of inattention increase the variability of responses in a perimetric type task. Redesign of the perimetric strategies to combat periods of inattention could significantly improve the reproducibility of visual field data.
Keywords: visual fields • electrophysiology: clinical • perimetry