ICA-based artefact removal and accelerated fMRI acquisition for improved resting state network imaging

Ludovica Griffanti; Gholamreza Salimi-Khorshidi; Christian F. Beckmann; Edward J. Auerbach; Gwenaëlle Douaud; Claire E. Sexton; Eniko Zsoldos; Klaus P. Ebmeier; Nicola Filippini; Clare E. Mackay; Steen Moeller; Junqian Xu; Essa Yacoub; Giuseppe Baselli; Kamil Ugurbil; Karla L. Miller; Stephen M. Smith

doi:10.1016/j.neuroimage.2014.03.034

ICA-based artefact removal and accelerated fMRI acquisition for improved resting state network imaging

Ludovica Griffanti, Gholamreza Salimi-Khorshidi, Christian F. Beckmann, Edward J. Auerbach, Gwenaëlle Douaud, Claire E. Sexton, Eniko Zsoldos, Klaus P. Ebmeier, Nicola Filippini, Clare E. Mackay, Steen Moeller, Junqian Xu, Essa Yacoub, Giuseppe Baselli, Kamil Ugurbil, Karla L. Miller, Stephen M. Smith

Radiology

Research output: Contribution to journal › Article › peer-review

819 Scopus citations

Abstract

The identification of resting state networks (RSNs) and the quantification of their functional connectivity in resting-state fMRI (rfMRI) are seriously hindered by the presence of artefacts, many of which overlap spatially or spectrally with RSNs. Moreover, recent developments in fMRI acquisition yield data with higher spatial and temporal resolutions, but may increase artefacts both spatially and/or temporally. Hence the correct identification and removal of non-neural fluctuations is crucial, especially in accelerated acquisitions. In this paper we investigate the effectiveness of three data-driven cleaning procedures, compare standard against higher (spatial and temporal) resolution accelerated fMRI acquisitions, and investigate the combined effect of different acquisitions and different cleanup approaches. We applied single-subject independent component analysis (ICA), followed by automatic component classification with FMRIB's ICA-based X-noiseifier (FIX) to identify artefactual components. We then compared two first-level (within-subject) cleaning approaches for removing those artefacts and motion-related fluctuations from the data. The effectiveness of the cleaning procedures was assessed using time series (amplitude and spectra), network matrix and spatial map analyses. For time series and network analyses we also tested the effect of a second-level cleaning (informed by group-level analysis). Comparing these approaches, the preferable balance between noise removal and signal loss was achieved by regressing out of the data the full space of motion-related fluctuations and only the unique variance of the artefactual ICA components. Using similar analyses, we also investigated the effects of different cleaning approaches on data from different acquisition sequences. With the optimal cleaning procedures, functional connectivity results from accelerated data were statistically comparable or significantly better than the standard (unaccelerated) acquisition, and, crucially, with higher spatial and temporal resolution. Moreover, we were able to perform higher dimensionality ICA decompositions with the accelerated data, which is very valuable for detailed network analyses.

Original language	English (US)
Pages (from-to)	232-247
Number of pages	16
Journal	NeuroImage
Volume	95
DOIs	https://doi.org/10.1016/j.neuroimage.2014.03.034
State	Published - Jul 15 2014

Bibliographical note

Funding Information:
Funding acknowledgements: NIH 1U54MH091657-01 P30-NS057091 P41-RR08079/EB015894 , UK MRC G1001354 (Whitehall II MRI Sub-study); The HDH Wills 1965 Charitable Trust (NF, AM). CES and CEM were supported by the National Institute for Health Research (NIHR) Oxford Biomedical Research Centre based at Oxford University Hospitals NHS Trust and University of Oxford. We thank: Charlotte Allan, Abda Mahmood, Amanda Pipkin, Anja Topiwala for recruiting and assessing participants; Radiographers (Caroline Young, Michael Sanders, Jon Campbell) for acquiring scans; Mika Kivimäki, Archana Singh-Manoux for giving access to Whitehall II data, and John R Geddes for applying for the funding.

Keywords

Artefact removal
Functional connectivity
Functional magnetic resonance imaging (fMRI)
Multiband acceleration
Resting-state

Access

10.1016/j.neuroimage.2014.03.034

OpenUrl availability

Full text

Cite this

Griffanti, L., Salimi-Khorshidi, G., Beckmann, C. F., Auerbach, E. J., Douaud, G., Sexton, C. E., Zsoldos, E., Ebmeier, K. P., Filippini, N., Mackay, C. E., Moeller, S., Xu, J., Yacoub, E., Baselli, G., Ugurbil, K., Miller, K. L., & Smith, S. M. (2014). ICA-based artefact removal and accelerated fMRI acquisition for improved resting state network imaging. NeuroImage, 95, 232-247. https://doi.org/10.1016/j.neuroimage.2014.03.034

Griffanti, L, Salimi-Khorshidi, G, Beckmann, CF, Auerbach, EJ, Douaud, G, Sexton, CE, Zsoldos, E, Ebmeier, KP, Filippini, N, Mackay, CE, Moeller, S, Xu, J, Yacoub, E, Baselli, G, Ugurbil, K, Miller, KL & Smith, SM 2014, 'ICA-based artefact removal and accelerated fMRI acquisition for improved resting state network imaging', NeuroImage, vol. 95, pp. 232-247. https://doi.org/10.1016/j.neuroimage.2014.03.034

@article{ff62fcd2167d42e5b75ffa814abd06d2,

title = "ICA-based artefact removal and accelerated fMRI acquisition for improved resting state network imaging",

abstract = "The identification of resting state networks (RSNs) and the quantification of their functional connectivity in resting-state fMRI (rfMRI) are seriously hindered by the presence of artefacts, many of which overlap spatially or spectrally with RSNs. Moreover, recent developments in fMRI acquisition yield data with higher spatial and temporal resolutions, but may increase artefacts both spatially and/or temporally. Hence the correct identification and removal of non-neural fluctuations is crucial, especially in accelerated acquisitions. In this paper we investigate the effectiveness of three data-driven cleaning procedures, compare standard against higher (spatial and temporal) resolution accelerated fMRI acquisitions, and investigate the combined effect of different acquisitions and different cleanup approaches. We applied single-subject independent component analysis (ICA), followed by automatic component classification with FMRIB's ICA-based X-noiseifier (FIX) to identify artefactual components. We then compared two first-level (within-subject) cleaning approaches for removing those artefacts and motion-related fluctuations from the data. The effectiveness of the cleaning procedures was assessed using time series (amplitude and spectra), network matrix and spatial map analyses. For time series and network analyses we also tested the effect of a second-level cleaning (informed by group-level analysis). Comparing these approaches, the preferable balance between noise removal and signal loss was achieved by regressing out of the data the full space of motion-related fluctuations and only the unique variance of the artefactual ICA components. Using similar analyses, we also investigated the effects of different cleaning approaches on data from different acquisition sequences. With the optimal cleaning procedures, functional connectivity results from accelerated data were statistically comparable or significantly better than the standard (unaccelerated) acquisition, and, crucially, with higher spatial and temporal resolution. Moreover, we were able to perform higher dimensionality ICA decompositions with the accelerated data, which is very valuable for detailed network analyses.",

keywords = "Artefact removal, Functional connectivity, Functional magnetic resonance imaging (fMRI), Multiband acceleration, Resting-state",

author = "Ludovica Griffanti and Gholamreza Salimi-Khorshidi and Beckmann, {Christian F.} and Auerbach, {Edward J.} and Gwena{\"e}lle Douaud and Sexton, {Claire E.} and Eniko Zsoldos and Ebmeier, {Klaus P.} and Nicola Filippini and Mackay, {Clare E.} and Steen Moeller and Junqian Xu and Essa Yacoub and Giuseppe Baselli and Kamil Ugurbil and Miller, {Karla L.} and Smith, {Stephen M.}",

note = "Funding Information: Funding acknowledgements: NIH 1U54MH091657-01 P30-NS057091 P41-RR08079/EB015894 , UK MRC G1001354 (Whitehall II MRI Sub-study); The HDH Wills 1965 Charitable Trust (NF, AM). CES and CEM were supported by the National Institute for Health Research (NIHR) Oxford Biomedical Research Centre based at Oxford University Hospitals NHS Trust and University of Oxford. We thank: Charlotte Allan, Abda Mahmood, Amanda Pipkin, Anja Topiwala for recruiting and assessing participants; Radiographers (Caroline Young, Michael Sanders, Jon Campbell) for acquiring scans; Mika Kivim{\"a}ki, Archana Singh-Manoux for giving access to Whitehall II data, and John R Geddes for applying for the funding. ",

year = "2014",

month = jul,

day = "15",

doi = "10.1016/j.neuroimage.2014.03.034",

language = "English (US)",

volume = "95",

pages = "232--247",

journal = "NeuroImage",

issn = "1053-8119",

publisher = "Academic Press Inc.",

}

TY - JOUR

T1 - ICA-based artefact removal and accelerated fMRI acquisition for improved resting state network imaging

AU - Griffanti, Ludovica

AU - Salimi-Khorshidi, Gholamreza

AU - Beckmann, Christian F.

AU - Auerbach, Edward J.

AU - Douaud, Gwenaëlle

AU - Sexton, Claire E.

AU - Zsoldos, Eniko

AU - Ebmeier, Klaus P.

AU - Filippini, Nicola

AU - Mackay, Clare E.

AU - Moeller, Steen

AU - Xu, Junqian

AU - Yacoub, Essa

AU - Baselli, Giuseppe

AU - Ugurbil, Kamil

AU - Miller, Karla L.

AU - Smith, Stephen M.

N1 - Funding Information: Funding acknowledgements: NIH 1U54MH091657-01 P30-NS057091 P41-RR08079/EB015894 , UK MRC G1001354 (Whitehall II MRI Sub-study); The HDH Wills 1965 Charitable Trust (NF, AM). CES and CEM were supported by the National Institute for Health Research (NIHR) Oxford Biomedical Research Centre based at Oxford University Hospitals NHS Trust and University of Oxford. We thank: Charlotte Allan, Abda Mahmood, Amanda Pipkin, Anja Topiwala for recruiting and assessing participants; Radiographers (Caroline Young, Michael Sanders, Jon Campbell) for acquiring scans; Mika Kivimäki, Archana Singh-Manoux for giving access to Whitehall II data, and John R Geddes for applying for the funding.

PY - 2014/7/15

Y1 - 2014/7/15

N2 - The identification of resting state networks (RSNs) and the quantification of their functional connectivity in resting-state fMRI (rfMRI) are seriously hindered by the presence of artefacts, many of which overlap spatially or spectrally with RSNs. Moreover, recent developments in fMRI acquisition yield data with higher spatial and temporal resolutions, but may increase artefacts both spatially and/or temporally. Hence the correct identification and removal of non-neural fluctuations is crucial, especially in accelerated acquisitions. In this paper we investigate the effectiveness of three data-driven cleaning procedures, compare standard against higher (spatial and temporal) resolution accelerated fMRI acquisitions, and investigate the combined effect of different acquisitions and different cleanup approaches. We applied single-subject independent component analysis (ICA), followed by automatic component classification with FMRIB's ICA-based X-noiseifier (FIX) to identify artefactual components. We then compared two first-level (within-subject) cleaning approaches for removing those artefacts and motion-related fluctuations from the data. The effectiveness of the cleaning procedures was assessed using time series (amplitude and spectra), network matrix and spatial map analyses. For time series and network analyses we also tested the effect of a second-level cleaning (informed by group-level analysis). Comparing these approaches, the preferable balance between noise removal and signal loss was achieved by regressing out of the data the full space of motion-related fluctuations and only the unique variance of the artefactual ICA components. Using similar analyses, we also investigated the effects of different cleaning approaches on data from different acquisition sequences. With the optimal cleaning procedures, functional connectivity results from accelerated data were statistically comparable or significantly better than the standard (unaccelerated) acquisition, and, crucially, with higher spatial and temporal resolution. Moreover, we were able to perform higher dimensionality ICA decompositions with the accelerated data, which is very valuable for detailed network analyses.

AB - The identification of resting state networks (RSNs) and the quantification of their functional connectivity in resting-state fMRI (rfMRI) are seriously hindered by the presence of artefacts, many of which overlap spatially or spectrally with RSNs. Moreover, recent developments in fMRI acquisition yield data with higher spatial and temporal resolutions, but may increase artefacts both spatially and/or temporally. Hence the correct identification and removal of non-neural fluctuations is crucial, especially in accelerated acquisitions. In this paper we investigate the effectiveness of three data-driven cleaning procedures, compare standard against higher (spatial and temporal) resolution accelerated fMRI acquisitions, and investigate the combined effect of different acquisitions and different cleanup approaches. We applied single-subject independent component analysis (ICA), followed by automatic component classification with FMRIB's ICA-based X-noiseifier (FIX) to identify artefactual components. We then compared two first-level (within-subject) cleaning approaches for removing those artefacts and motion-related fluctuations from the data. The effectiveness of the cleaning procedures was assessed using time series (amplitude and spectra), network matrix and spatial map analyses. For time series and network analyses we also tested the effect of a second-level cleaning (informed by group-level analysis). Comparing these approaches, the preferable balance between noise removal and signal loss was achieved by regressing out of the data the full space of motion-related fluctuations and only the unique variance of the artefactual ICA components. Using similar analyses, we also investigated the effects of different cleaning approaches on data from different acquisition sequences. With the optimal cleaning procedures, functional connectivity results from accelerated data were statistically comparable or significantly better than the standard (unaccelerated) acquisition, and, crucially, with higher spatial and temporal resolution. Moreover, we were able to perform higher dimensionality ICA decompositions with the accelerated data, which is very valuable for detailed network analyses.

KW - Artefact removal

KW - Functional connectivity

KW - Functional magnetic resonance imaging (fMRI)

KW - Multiband acceleration

KW - Resting-state

UR - http://www.scopus.com/inward/record.url?scp=84899916725&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84899916725&partnerID=8YFLogxK

U2 - 10.1016/j.neuroimage.2014.03.034

DO - 10.1016/j.neuroimage.2014.03.034

M3 - Article

C2 - 24657355

AN - SCOPUS:84899916725

SN - 1053-8119

VL - 95

SP - 232

EP - 247

JO - NeuroImage

JF - NeuroImage

ER -

ICA-based artefact removal and accelerated fMRI acquisition for improved resting state network imaging

Abstract

Bibliographical note

Keywords

Access

OpenUrl availability

Other files and links

Fingerprint

Cite this