Adaptive Smoothing as Inference Strategy

More Specificity for Unequally Sized or Neighbouring Regions

Marijke Welvaert, Karsten Tabelow, Ruth Seurinck, Yves Rosseel

Research output: Contribution to journalArticle

1 Citation (Scopus)

Abstract

Although spatial smoothing of fMRI data can serve multiple purposes, increasing the sensitivity of activation detection is probably its greatest benefit. However, this increased detection power comes with a loss of specificity when non-adaptive smoothing (i.e. the standard in most software packages) is used. Simulation studies and analysis of experimental data was performed using the R packages neuRosim and fmri. In these studies, we systematically investigated the effect of spatial smoothing on the power and number of false positives in two particular cases that are often encountered in fMRI research: (1) Single condition activation detection for regions that differ in size, and (2) multiple condition activation detection for neighbouring regions. Our results demonstrate that adaptive smoothing is superior in both cases because less false positives are introduced by the spatial smoothing process compared to standard Gaussian smoothing or FDR inference of unsmoothed data.
Original languageEnglish
Pages (from-to)435-445
Number of pages9
JournalNeuroInformatics
Volume11
Issue number4
DOIs
Publication statusPublished - Oct 2013
Externally publishedYes

Cite this

Welvaert, Marijke ; Tabelow, Karsten ; Seurinck, Ruth ; Rosseel, Yves. / Adaptive Smoothing as Inference Strategy : More Specificity for Unequally Sized or Neighbouring Regions. In: NeuroInformatics. 2013 ; Vol. 11, No. 4. pp. 435-445.
@article{6acc396c7c9f4b53969d1488c4193f9c,
title = "Adaptive Smoothing as Inference Strategy: More Specificity for Unequally Sized or Neighbouring Regions",
abstract = "Although spatial smoothing of fMRI data can serve multiple purposes, increasing the sensitivity of activation detection is probably its greatest benefit. However, this increased detection power comes with a loss of specificity when non-adaptive smoothing (i.e. the standard in most software packages) is used. Simulation studies and analysis of experimental data was performed using the R packages neuRosim and fmri. In these studies, we systematically investigated the effect of spatial smoothing on the power and number of false positives in two particular cases that are often encountered in fMRI research: (1) Single condition activation detection for regions that differ in size, and (2) multiple condition activation detection for neighbouring regions. Our results demonstrate that adaptive smoothing is superior in both cases because less false positives are introduced by the spatial smoothing process compared to standard Gaussian smoothing or FDR inference of unsmoothed data.",
keywords = "False positive rate, fMRI, Gaussian smoothing, Power, Structural adaptive segmentation",
author = "Marijke Welvaert and Karsten Tabelow and Ruth Seurinck and Yves Rosseel",
year = "2013",
month = "10",
doi = "10.1007/s12021-013-9196-z",
language = "English",
volume = "11",
pages = "435--445",
journal = "NeuroInformatics",
issn = "1539-2791",
publisher = "Humana Press",
number = "4",

}

Adaptive Smoothing as Inference Strategy : More Specificity for Unequally Sized or Neighbouring Regions. / Welvaert, Marijke; Tabelow, Karsten; Seurinck, Ruth; Rosseel, Yves.

In: NeuroInformatics, Vol. 11, No. 4, 10.2013, p. 435-445.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Adaptive Smoothing as Inference Strategy

T2 - More Specificity for Unequally Sized or Neighbouring Regions

AU - Welvaert, Marijke

AU - Tabelow, Karsten

AU - Seurinck, Ruth

AU - Rosseel, Yves

PY - 2013/10

Y1 - 2013/10

N2 - Although spatial smoothing of fMRI data can serve multiple purposes, increasing the sensitivity of activation detection is probably its greatest benefit. However, this increased detection power comes with a loss of specificity when non-adaptive smoothing (i.e. the standard in most software packages) is used. Simulation studies and analysis of experimental data was performed using the R packages neuRosim and fmri. In these studies, we systematically investigated the effect of spatial smoothing on the power and number of false positives in two particular cases that are often encountered in fMRI research: (1) Single condition activation detection for regions that differ in size, and (2) multiple condition activation detection for neighbouring regions. Our results demonstrate that adaptive smoothing is superior in both cases because less false positives are introduced by the spatial smoothing process compared to standard Gaussian smoothing or FDR inference of unsmoothed data.

AB - Although spatial smoothing of fMRI data can serve multiple purposes, increasing the sensitivity of activation detection is probably its greatest benefit. However, this increased detection power comes with a loss of specificity when non-adaptive smoothing (i.e. the standard in most software packages) is used. Simulation studies and analysis of experimental data was performed using the R packages neuRosim and fmri. In these studies, we systematically investigated the effect of spatial smoothing on the power and number of false positives in two particular cases that are often encountered in fMRI research: (1) Single condition activation detection for regions that differ in size, and (2) multiple condition activation detection for neighbouring regions. Our results demonstrate that adaptive smoothing is superior in both cases because less false positives are introduced by the spatial smoothing process compared to standard Gaussian smoothing or FDR inference of unsmoothed data.

KW - False positive rate

KW - fMRI

KW - Gaussian smoothing

KW - Power

KW - Structural adaptive segmentation

U2 - 10.1007/s12021-013-9196-z

DO - 10.1007/s12021-013-9196-z

M3 - Article

VL - 11

SP - 435

EP - 445

JO - NeuroInformatics

JF - NeuroInformatics

SN - 1539-2791

IS - 4

ER -