Graduation Semester and Year

2019

Language

English

Document Type

Thesis

Degree Name

Master of Science in Biomedical Engineering

Department

Bioengineering

First Advisor

Hanli Liu

Abstract

Neuroeconomics is an emerging field integrating economic theories with neuroscience to enhance the understanding on how humans make decisions under different cost-effective conditions. Understanding the brain network associated with different cognitive tasks, to investigate the basic neural processes that underlie complex higher-order cognitive operations, has become an important research topic in behavioral neuroscience. Newsvendor problem plays a vital role among all prevalent concepts. This study is designed based on Newsvendor problem incorporating 40 trials for each subject at a randomly chosen low-profit or high-profit margin treatments to investigate how electrophysiological signals in the human brain are differentiated under the two treatments using 64-channel electroencephalography. The electrophysiological data were collected from 13 subjects while they were making decisions under randomly assigned two treatment levels with a MATLAB-based newsvendor game. Power density analysis of EEG signals was performed in 5 frequency components of EEG, which were delta (1–4 Hz), theta (4 –8 Hz), alpha (8–13 Hz), beta (13–30 Hz), and gamma (30–70 Hz) for three experimental phases, namely: baseline, decision iii making, and feedback. Root-Mean-Square (RMS) values of power calculation from each frequency band of each electrode was calculated using MATLAB; then 64-channel RMS magnitudes were averaged over all the participants to generate group-level topographical maps for all five frequency bands and throughout the three different phases. The group-level topographies across all five bands show significant activations in several major scalp regions during high- and low-margin task phases. Paired-sample t-tests with falsediscovery-rate correction (p<0.05) were conducted to identify significant electrodes across different treatments for multi-variable comparisons. Five clusters were identified from the 64- channel scalp region based on the common electrodes that shows statistical significance in one of the frequency bands. The cluster-based analysis indicated that significant activation by the newsvendor decision making occurred in the dorsolateral prefrontal cortex at four brain rhythm bands, excluding the gamma frequency. Also, power densities in alpha and theta bands shows opposite activation trends during the decision-making and feedback phase. Previous studies reported that the theta band plays a key role in memory retrieval and decision making. Our findings showed that the theta activation was significantly observed in all five clusters during the newsvendor decision-making phase. On the other hand, this specific decision making did not cause any change in beta band power. However, after splitting these data into low- and high-margin treatment conditions, it showed significant deactivation in the right dorsolateral prefrontal cortex during the decision-making phase under the low-margin treatment, reflecting mental stress or anxiety, as expected.

Keywords

Decision making study, Electroencephalography, RMS-based power calculation, Cluster-based analysis

Disciplines

Biomedical Engineering and Bioengineering | Engineering

Comments

Degree granted by The University of Texas at Arlington

Download

Share

COinS