RANDOM PROCESS MODEL FOR EVOKED EEG RESPONSES.
Abstract
This research has been concerned with proposing a mathematical model for relating the response of the brain to certain stimulus fields and then testing the usefulness of the model with experimental data. The model utilizes certain fundamental concepts of statistical theory of communication and control systems to describe the brain's processing of information in terms of operations on the input signal. The electrical activity of the brain is represented as a lumped-parameter random process. That is, the EEG recorded at each electrode represents the brain's activity within some neighborhood of the electrode. These signals contain two independent components: (1) the response of the brain to the input stimulus, and (2) the normal ongoing activity of the brain. For a linear, time-invariant system, the transmission operators have been designated in the time domain by the impulse response. Stipulating that the impulse response is a random process has led to several interesting input-output relationships for the model in terms of mathematical expectations and time averages. In particular, an averaging process has been formulated for relating the ensemble averages associated with the model to time averages of the measured signal. That is, evoked responses of the brain are used to compute the expected impulse responses and expected transfer functions of the random process. Changes in the characteristics of the evoked potentials and ongoing activity which are due to changes in the subject's psycho-neurophysiological state rather than the basic randomness of the system, are accounted for in the ensemble averages by an additional parameter in the system.
Document Details
- Document Type
- Technical Report
- Publication Date
- Jan 28, 1966
- Accession Number
- AD0489075
Entities
People
- Ashley J. Welch
Organizations
- University of Texas at Austin