City Research Online

Spectral analysis of fMRI signal and noise

Chen, C-C. and Tyler, C. W. (2008). Spectral analysis of fMRI signal and noise. In: Onozuka, M. and Chen-Tung, Y. (Eds.), Novel Trends in Brain Science. (pp. 63-76). Japan: Springer Japan. ISBN 9784431732419

Abstract

We analyzed the noise in functional magnetic resonance imaging (fMRI) scans of the human brain during rest. The noise spectrum in the cortex is well fitted by a model consisting of two additive components: flat-spectrum noise that is uniform throughout the MRI image and frequency-dependent biological noise that is localized to the neural tissue and declines from low to high temporal frequencies. We show that the frequency-dependent component is well fitted by the f -p model with 0 < p < 1 throughout the measured frequency range. The parameters of the model indicate that the characteristic noise is not attributable to the temporal filtering of the hemodynamic response but is an inherent property of the blood oxygenation level-dependent (BOLD) signal. We then analyzed the power spectrum of the BOLD signal for various cognitive tasks. The signal-to-noise ratio of a typical fMRI experiment peaks at around 0.04 Hz.

Publication Type: Book Section
Publisher Keywords: Blood oxygenation level-dependent (BOLD); Cortex; Rest scan; Stochastic process
Subjects: R Medicine > RC Internal medicine > RC0321 Neuroscience. Biological psychiatry. Neuropsychiatry
Departments: School of Health Sciences > Optometry & Visual Science
URI: http://openaccess.city.ac.uk/id/eprint/7056
[img]
Preview
Text - Accepted Version
Download (10MB) | Preview

Export

Downloads

Downloads per month over past year

View more statistics

Actions (login required)

Admin Login Admin Login