Multiple time scale complexity analysis of resting state FMRI - PubMed (original) (raw)

Multiple time scale complexity analysis of resting state FMRI

Robert X Smith et al. Brain Imaging Behav. 2014 Jun.

Abstract

The present study explored multi-scale entropy (MSE) analysis to investigate the entropy of resting state fMRI signals across multiple time scales. MSE analysis was developed to distinguish random noise from complex signals since the entropy of the former decreases with longer time scales while the latter signal maintains its entropy due to a "self-resemblance" across time scales. A long resting state BOLD fMRI (rs-fMRI) scan with 1000 data points was performed on five healthy young volunteers to investigate the spatial and temporal characteristics of entropy across multiple time scales. A shorter rs-fMRI scan with 240 data points was performed on a cohort of subjects consisting of healthy young (age 23 ± 2 years, n = 8) and aged volunteers (age 66 ± 3 years, n = 8) to investigate the effect of healthy aging on the entropy of rs-fMRI. The results showed that MSE of gray matter, rather than white matter, resembles closely that of f (-1) noise over multiple time scales. By filtering out high frequency random fluctuations, MSE analysis is able to reveal enhanced contrast in entropy between gray and white matter, as well as between age groups at longer time scales. Our data support the use of MSE analysis as a validation metric for quantifying the complexity of rs-fMRI signals.

PubMed Disclaimer

Conflict of interest statement

Disclosure statement: The authors have no conflicts of interest to disclose.

Figures

Fig. 1

Fig. 1

Illustration of the calculation of SampEn of rs-fMRI at the original time scale of 1 with a pattern length of m and a threshold of r. Here the red-blue patterns will be considered a match for m=2, but the red-blue-green patterns will not be counted for m=3, indicating an irregular process.

Fig. 2

Fig. 2

Group average gray matter and white matter SampEn over multiple time scales, before and after motion correction. SampEn of Gaussian-distributed uncorrelated (white) and correlated (pink _f_−1) noise are plotted for comparison.

Fig. 3

Fig. 3

Test for the effect of spontaneous fluctuations on entropy. Comparison of average SampEn for gray matter and white matter for a single volunteer over multiple time scales at TE = 10 ms and TE = 30 ms. The latter is the approximate time for optimal BOLD contrast. The inset shows the percent change in entropy as TE is increased from 10 ms to 30 ms. The displayed error bars are approximately the size of the symbols.

Fig. 4

Fig. 4

Group mean SampEn images for three slices, Z=36, 48, and 59 (bottom, middle, and top, respectively) in MNI space, of five volunteers at scales 1, 4, 7, and 10.

Fig. 5

Fig. 5

Average gray matter entropy for 8 young volunteers (age 23±2 years) and 8 aged volunteers (age 66±3 years). Plotted error bars are four standard errors of the respective means, p<10−4, and approximately the size of the symbols.

Fig. 6

Fig. 6

Increased regional MSE in young subjects. Images show results of multivariate two sample t test comparing healthy young subjects versus healthy aged subjects. Only clusters with 18 or more activated voxels (p<0.05, corrected) are shown. Decreases in MSE in older subjects are seen in regions associated with the default mode network: middle temporal gyrus - MTG, anterior cingulate gyrus - ACG, left and right angular gyrus - AG, middle and superior medial frontal cortex - MFG, SFGm. Significant decreases are also seen in the thalamus - THAL, caudate - CD, the lingual gyrus - LING, the hippocampus -HIPP, the supramarginal gyrus - SMG, and the superior temporal gyrus - STG. The numbers to the top left of each image refer to the z coordinate in MNI space.

References

    1. Abasolo D, Hornero R, Espino P, Poza J, Sanchez C, de la Rosa R. Analysis of regularity in the eeg background activity of alzheimer's disease patients with approximate entropy. Clinical neurophysiology. 2005;116:1826–1834. -PubMed
    1. Biswal B, Mennes M, Zuo X, Gohel S, Kelly C, Smith S, et al. Toward discovery science of human brain function. Proceedings of the National Academy of Sciences of the United States of America. 2010;107:4734–4739. -PMC -PubMed
    1. Bullmore E, Fadili J, Maxim V, Sendur L, Whitcher B, Suckling J, et al. Wavelets and functional magnetic resonance imaging of the human brain. NeuroImage. 2004;23 -PubMed
    1. Chang C, Glover G. Time-frequency dynamics of resting-state brain connectivity measured with fmri. NeuroImage. 2010;50:81–98. -PMC -PubMed
    1. Ciuciu P, Varoquaux G, Abry P, Sadaghiani S, Kleinschmidt A. Scale-free and multifractal properties of fmri signals during rest and task. Frontiers in Physiology. 2012;3(186) -PMC -PubMed

Publication types

MeSH terms

Substances

Grants and funding

LinkOut - more resources