Aarhus University Seal / Aarhus Universitets segl

Torben Ellegaard Lund

Sampling rate dependence of correlation at long time lags in BOLD fMRI measurements on humans and gel phantoms

Research output: Contribution to journal/Conference contribution in journal/Contribution to newspaperJournal articleResearchpeer-review

Standard

Sampling rate dependence of correlation at long time lags in BOLD fMRI measurements on humans and gel phantoms. / Mikkelsen, Kaare B; Lund, Torben Ellegaard.

In: Frontiers in Physiology, Vol. 4, 2013, p. 106.

Research output: Contribution to journal/Conference contribution in journal/Contribution to newspaperJournal articleResearchpeer-review

Harvard

APA

CBE

MLA

Vancouver

Author

Bibtex

@article{99e0826bcbdb4a5bb529e8758e330104,
title = "Sampling rate dependence of correlation at long time lags in BOLD fMRI measurements on humans and gel phantoms",
abstract = "The aim of this study is to investigate the effects of sampling rate on Hurst exponents derived from Blood Oxygenation Level Dependent functional Magnetic Resonance Imaging (BOLD fMRI) resting state time series. fMRI measurements were performed on 2 human subjects and a selection of gel phantoms. From these, Hurst exponents were calculated. It was found that low sampling rates induced non-trivial exponents at sharp material transitions, and that Hurst exponents of human measurements had a strong TR-dependence. The findings are compared to theoretical considerations regarding the fractional Gaussian noise model and resampling, and it is found that the implications are problematic. This result should have a direct influence on the way future studies of low-frequency variation in BOLD fMRI data are conducted, especially if the fractional Gaussian noise model is considered. We recommend either using a different model (examples of such are referenced in the conclusion), or standardizing experimental procedures along an optimal sampling rate.",
author = "Mikkelsen, {Kaare B} and Lund, {Torben Ellegaard}",
year = "2013",
doi = "10.3389/fphys.2013.00106",
language = "English",
volume = "4",
pages = "106",
journal = "Frontiers in Physiology",
issn = "1664-042X",
publisher = "Frontiers Media S.A",

}

RIS

TY - JOUR

T1 - Sampling rate dependence of correlation at long time lags in BOLD fMRI measurements on humans and gel phantoms

AU - Mikkelsen, Kaare B

AU - Lund, Torben Ellegaard

PY - 2013

Y1 - 2013

N2 - The aim of this study is to investigate the effects of sampling rate on Hurst exponents derived from Blood Oxygenation Level Dependent functional Magnetic Resonance Imaging (BOLD fMRI) resting state time series. fMRI measurements were performed on 2 human subjects and a selection of gel phantoms. From these, Hurst exponents were calculated. It was found that low sampling rates induced non-trivial exponents at sharp material transitions, and that Hurst exponents of human measurements had a strong TR-dependence. The findings are compared to theoretical considerations regarding the fractional Gaussian noise model and resampling, and it is found that the implications are problematic. This result should have a direct influence on the way future studies of low-frequency variation in BOLD fMRI data are conducted, especially if the fractional Gaussian noise model is considered. We recommend either using a different model (examples of such are referenced in the conclusion), or standardizing experimental procedures along an optimal sampling rate.

AB - The aim of this study is to investigate the effects of sampling rate on Hurst exponents derived from Blood Oxygenation Level Dependent functional Magnetic Resonance Imaging (BOLD fMRI) resting state time series. fMRI measurements were performed on 2 human subjects and a selection of gel phantoms. From these, Hurst exponents were calculated. It was found that low sampling rates induced non-trivial exponents at sharp material transitions, and that Hurst exponents of human measurements had a strong TR-dependence. The findings are compared to theoretical considerations regarding the fractional Gaussian noise model and resampling, and it is found that the implications are problematic. This result should have a direct influence on the way future studies of low-frequency variation in BOLD fMRI data are conducted, especially if the fractional Gaussian noise model is considered. We recommend either using a different model (examples of such are referenced in the conclusion), or standardizing experimental procedures along an optimal sampling rate.

U2 - 10.3389/fphys.2013.00106

DO - 10.3389/fphys.2013.00106

M3 - Journal article

C2 - 23730289

VL - 4

SP - 106

JO - Frontiers in Physiology

JF - Frontiers in Physiology

SN - 1664-042X

ER -