Analysis of Single Event Evoked Potentials.

Abstract

The purpose of this exploratory research program was to demonstrate the feasibility of on-line classification of brain waves associated with specific visual stimulus events. A number of classification and discrimination techniques were tested and compared. It was found that by utilizing 4 electrodes placed on the scalp and measuring the voltage amplitude at no more than six time points, a classification accuracy of over 90% was achieved for choosing between the four stimuli: blank field; full field checkerboard; upper half field checkerboard, and lower half field checkerboard. Other types of visual stimuli were utilized and their brain responses classified using techniques developed for the earlier set of tests. Visual stimuli investigated included: stimulation of the upper and lower half visual fields and the right and left visual fields with a checkerboard pattern. Stimulation of the four visual quadrants and the full visual field with a checkerboard pattern. In addition, the effects of focusing or defocusing a letter on the brain potentials were examined. Classification results obtained with these experiments were always in the high 80% category or low 90%. The results obtained were very encouraging and suggest that these techniques may be suitable for practical application.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Nov 01, 1979
Accession Number
ADA080896

Entities

People

  • Clare D. Mcgillem
  • Jorge I. Aunon

Organizations

  • Purdue University

Tags

Communities of Interest

  • Biomedical
  • Materials and Manufacturing Processes

DTIC Thesaurus Topics

  • Accuracy
  • Air Force
  • Air Force Facilities
  • Bioengineering
  • Biomedical Engineering
  • Computer Programs
  • Data Processing
  • Electrical Engineering
  • Electrophysiological Phenomena
  • Feature Selection
  • Frequency
  • Information Processing
  • Information Science
  • Medical Laboratories
  • Psychology
  • Recording Systems
  • Signal Processing

Readers

  • Hydrologic Risk Analysis and Mitigation.
  • Image Processing and Computer Vision.
  • Neuroscience