Central-Tendency Estimation and Nearest-Estimate Classification of Multi-Channel Evoked Potentials.

Author

Phani Srinivas Prasad Yarlagadda, Southern Illinois University CarbondaleFollow

Date of Award

12-2009

Degree Name

Master of Science

Department

Electrical and Computer Engineering

First Advisor

Gupta, Lalit

Abstract

By modeling evoked potentials (EPs) as random vectors in which the EP samples are random variables, a unified strategy is introduced to determine multivariate central-tendency estimates such as the arithmetic mean, geometric mean, harmonic mean, median, tri-mean, trimmed-mean, and the Winsorized mean. Additionally, a generalized strategy is introduced to develop minimum-distance classifiers based on central-tendency estimates. Furthermore, procedures are developed to fuse the decisions of the resulting nearest-estimate classifiers for single-channel heterogeneous, multi-channel homogeneous, and multichannel heterogeneous-homogenous EP classification. The central-tendency estimates of real EPs are compared and it is shown that although the operations to compute the vector central-tendency estimates can be quite different, the EP estimates are similar with respect to their overall waveform shapes and peak latencies. It is also shown that by fusing homogeneous nearest-estimate classifier decisions across multiple channels, the classification accuracy can be improved significantly when compared with the accuracies of individual channel classifiers.