Neural Network Classification of EEG Using Chaotic Preprocessing and Phase Space Reconstruction

Abstract

Researchers have long focussed on using physiologic measures as an indicator of mental workload. Research with transient and steady state evoked potential electroencephalograph (EEG) signals has provided the basis for further study in mental state estimation. Studies involving sum-of-sines steady state evoked potentials have demonstrated a correlation between spectral changes and changing cognitive workload. It has also been found that subjects can learn to control their responses to steady state visual stimulus, provided near-real-time performance information was fed back to them in such a way as to close the loop encompassing the subjects and the stimulus. With the emergence of new sciences such as Artificial Neural Systems and Chaotic theory, the possibility of achieving a rudimentary form of automatic cognitive state estimation or 'Cognitive Mode Mapping' has presented itself. Using these powerful analysis tools, the authors are developing a system that analyzes and classifies EEG data from four sites of a subject's brain. The subjects produce this data while performing five selected cognitive tasks. The objective of the Cognitive Mode Mapping system is to identify the tasks based on salient features embedded in the raw EEG signals.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Apr 01, 1991
Accession Number
ADA279098

Entities

People

  • Craig W. Downey
  • David F. Ingle
  • David Tumey
  • John H. Schnurer
  • Paul E. Morton

Tags

Communities of Interest

  • Materials and Manufacturing Processes
  • Space

DTIC Thesaurus Topics

  • Air Force
  • Air Force Facilities
  • Biomedical Research
  • Brain
  • Classification
  • Cognitive Workload
  • Electrophysiological Phenomena
  • Government Procurement
  • Governments
  • Human Factors Engineering
  • Neural Networks
  • Pattern Recognition
  • Preprocessing
  • Steady State
  • Technical Information Centers
  • Test Sets
  • Workload

Readers

  • Brain and Cognitive Science; Experimental Psychology; Cognitive Neuroscience
  • Control Systems Engineering.
  • Neural Network Machine Learning.

Technology Areas

  • AI & ML
  • AI & ML - Bayesian Inference
  • AI & ML - Neural Networks
  • Space