Electro-Encephalogram Based Adaptive Estimation of Magneto-Encephalogram Signals
Abstract
Adaptive signal processing techniques were used to filter out unwanted background noise from the evoked response field signals obtained from magneto-encephalogram measurements. A model of the evoked response field signals was first developed to test the adaptive algorithm in an environment corrupted by white gaussian noise. Several modeling experiments verified the feasibility of adaptive filtering using an enhancement design with a correlated signal representing the evoked potential response obtained from electro-enchephalogram measurements. The experimental results showed that signal estimation is improved by a strong correlation between the evoke response field and evoke response potential. Following the modeling experiments, filtering of actual evoked responses was attempted. To obtain the evoked field data, an audio or visual stimulus was provided to a test subject located inside a shielded chamber. Time sequenced electro-encephalogram and magneto-encephalogram signals were recorded for later processing using an adaptive filter based on the least-mean-square algorithm. Accuracy of the filtered human data could not be quantified due to a lack of a priori knowledge of the exact signals before filtering. Comparisons of filtered responses with ensemble averaged responses of up to 70 signals showed waveform similarities. Keywords: Computerized simulation. Theses.
Document Details
- Document Type
- Technical Report
- Publication Date
- Dec 01, 1988
- Accession Number
- ADA202662
Entities
People
- Roger A. Wood
Organizations
- Air Force Institute of Technology