EEG-based cognitive load estimation

Publication Year:
2013
Usage 175
Downloads 142
Abstract Views 33
Repository URL:
https://digitalcommons.ohsu.edu/etd/888
DOI:
10.6083/m4833q22
Author(s):
Quinn, Max
Publisher(s):
Oregon Health & Science University
Tags:
Electroencephalography; Machine learning; Signal processing; Aphasia; Cognition; Artificial Intelligence; Signal Processing; Computer-Assisted; Signal Processing, Computer-Assisted
thesis / dissertation description
Understanding and monitoring cognitive processes is a difficult task; made more so by the apparently complex underlying resource networks that make up cognition, as well as by confounds such as variability in the effort expended by subjects while performing cognitive tasks. In this paper, we describe an electroencephalogram (EEG) and machine learning-based approach to estimating the cognitive effort exerted by a subject while performing a task. We describe the components of an EEG processing pipeline and a machine learning system, the challenges associated with employing these systems, and consider several implementation options. We contribute a novel method for separating ocular artifacts from cortical activity that represents a possible improvement upon existing techniques. In addition, we have investigated a number of alternative approaches to classification and found that these perform in a similar manner as those in prior research. Although these approaches did not produce desir