Characterization of acoustic imaging noise-induced hemodynamic response as a function of noise intensity in human auditory cortex fMRI at 3.0T

Minseok Kwon, Purdue University

Abstract

Acoustic imaging noise (AIN) associated with fast gradient switching during image acquisition is a confounding factor in auditory fMRI experiments. AIN interferes with the perception of desired stimuli as well as inducing hemodynamic responses that distort the responses to the desired stimuli. Moreover, non-linear behavior of the HDR to stimuli of varying intensities and durations make analysis of fMRI data more problematic. Therefore, characterization of the pure HDRs induced by AIN as a function of intensity of AIN may improve the analysis of responses to the desired auditory stimuli. In this study, we conducted two fMRI sessions on each of 11 subjects in order to determine the HDRs induced by AIN in the context of amplitude and spatial extents. Two AIN stimuli of varying duration were employed (1ping – 0.1 s duration, and 10 pings – 1 s duration), and two AIN intensities were evaluated by using earplug/earmuff combination and earmuff alone at 3.0T fMRI. Estimated HDRs to each stimuli duration and intensity were obtained by averaging across voxels within the region of interest (Heschl's gyrus), across runs, and across subjects. T-tests comparing the HDRs induced by stimuli of varying intensities show that there was not a statistically significant difference in HDRs. Moreover, all modeled HDRs using double gamma variate function based on estimated HDRs illustrate similar time-course. From the comparison of modeled HDRs to 1 ping with those to 10 pings, increase in amplitude and distribution was observed in a long ping duration. These results suggest that difference in acoustic imaging noise (15dB) we provided has little effect on waveform of HDRs. In statistical activation maps for 1 ping duration, AIN induced activation on primary auditory cortex as well as secondary auditory cortex, and spatial extents of activation between different AIN intensities were similar. For 10 pings AIN, earplug and earmuff combination made spatial extent of activation more concentrated on primary and secondary auditory cortex, which enable the investigation of distortion by AIN to focus on the auditory cortex. Therefore, large attenuation leads to reduced spatial extent of activation, especially for long AIN. Moreover, AIN at longer duration produced more activity which may generate more distortions on HDRs to desired stimulus.

Degree

M.S.E.C.E.

Advisors

Talavage, Purdue University.

Subject Area

Biomedical engineering|Electrical engineering

Off-Campus Purdue Users:
To access this dissertation, please log in to our
proxy server
.

Share

COinS