Graduation Year


Document Type




Degree Granting Department

Chemical and Biomedical Engineering

Major Professor

David A. Eddins, Ph.D.

Committee Member

Ann C. Eddins, Ph.D.

Committee Member

Joseph P. Walton, Ph.D


behavior, cortical distribution, late potentials, laterality, young adults


Natural sounds are characterized by the distribution of acoustic power over different frequency regions and/or time. This is termed spectral, temporal or spectro-temporal modulation. The auditory system is equipped with banks of filters tuned to different spectral, temporal and spectro-temporal modulation frequencies (SM, TM, STM). The sensitivity of the peripheral system to these modulations can be measured by undertaking a linear systems approach. In addition to understanding the psychophysical sensitivity, studying the neural patterns of their processing is also critical. The current study is an attempt to understand the relationship between the behavioral and neural correlates of spectral, temporal and spectro-temporal processing in ten normal hearing subjects (age range 21-27 years; mean = 23.7 years). In the behavioral experiment, sensitivity to SM, TM and STM frequencies was estimated using a 3-interval, 3-alternative, forced-choice paradigm with a 3-down-1-up tracking algorithm. In the electrophysiological experiment, Electroencephalographs (EEGs) were recorded in a change-detection paradigm in response to the same set of modulation stimuli used in the behavioral experiment presented at 20 dB sensation level (SL). The EEG data were analyzed to determine the global field power and latencies of the N1and P2 components and the amplitude of the N1-P2 complex. Although an overall parietal dominance was observed for all of the components, the N1-P2 complex was strongly lateralized to the right hemisphere in the frontal region, but the hemispheric asymmetry decreased at central and parietal regions. A highly significant but weak to moderate negative correlation between individual behavioral thresholds and N1-P2 amplitudes was observed, and this relationship also was observed when behavioral spectro-temporal transfer functions and N1-P2 amplitude transfer functions were examined together. Thus the current project reveals that a relationship exists between the behavioral measures and neural correlates and gives us hope to work towards establishing this relationship.