Asymmetric bias in user guided segmentations of brain structures

Eric Maltbie, Kshamta Bhatt, Beatriz Paniagua, Rachel G. Smith, Michael M. Graves, Matthew W. Mosconi, Sarah Peterson, Scott White, Joseph Blocher, Mohammed El-Sayed, Heather C. Hazlett, Martin A. Styner

Research output: Contribution to journalArticle

31 Citations (Scopus)

Abstract

Brain morphometric studies often incorporate comparative hemispheric asymmetry analyses of segmented brain structures. In this work, we present evidence that common user guided structural segmentation techniques exhibit strong left-right asymmetric biases and thus fundamentally influence any left-right asymmetry analyses. In this study, MRI scans from ten pediatric subjects were employed for studying segmentations of amygdala, globus pallidus, putamen, caudate, and lateral ventricle. Additionally, two pediatric and three adult scans were used for studying hippocampus segmentation. Segmentations of the sub-cortical structures were performed by skilled raters using standard manual and semi-automated methods. The left-right mirrored versions of each image were included in the data and segmented in a random order to assess potential left-right asymmetric bias. Using shape analysis we further assessed whether the asymmetric bias is consistent across subjects and raters with the focus on the hippocampus. The user guided segmentation techniques on the sub-cortical structures exhibited left-right asymmetric volume bias with the hippocampus displaying the most significant asymmetry values (p << 0.01). The hippocampal shape analysis revealed the bias to be strongest on the lateral side of the body and medial side of the head and tail. The origin of this asymmetric bias is considered to be based in laterality of visual perception; therefore segmentations with any degree of user interaction contain an asymmetric bias. The aim of our study is to raise awareness in the neuroimaging community regarding the presence of the asymmetric bias and its influence on any left-right hemispheric analyses. We also recommend reexamining previous research results in the light of this new finding.

Original languageEnglish (US)
Pages (from-to)1315-1323
Number of pages9
JournalNeuroImage
Volume59
Issue number2
DOIs
StatePublished - Jan 16 2012

Fingerprint

Hippocampus
Brain
Pediatrics
Visual Perception
Globus Pallidus
Lateral Ventricles
Putamen
Amygdala
Neuroimaging
Head
Magnetic Resonance Imaging
Research

Keywords

  • Asymmetry analysis
  • Manual segmentation
  • Rater bias
  • Subcortical structures neuroimaging

ASJC Scopus subject areas

  • Cognitive Neuroscience
  • Neurology

Cite this

Maltbie, E., Bhatt, K., Paniagua, B., Smith, R. G., Graves, M. M., Mosconi, M. W., ... Styner, M. A. (2012). Asymmetric bias in user guided segmentations of brain structures. NeuroImage, 59(2), 1315-1323. https://doi.org/10.1016/j.neuroimage.2011.08.025

Asymmetric bias in user guided segmentations of brain structures. / Maltbie, Eric; Bhatt, Kshamta; Paniagua, Beatriz; Smith, Rachel G.; Graves, Michael M.; Mosconi, Matthew W.; Peterson, Sarah; White, Scott; Blocher, Joseph; El-Sayed, Mohammed; Hazlett, Heather C.; Styner, Martin A.

In: NeuroImage, Vol. 59, No. 2, 16.01.2012, p. 1315-1323.

Research output: Contribution to journalArticle

Maltbie, E, Bhatt, K, Paniagua, B, Smith, RG, Graves, MM, Mosconi, MW, Peterson, S, White, S, Blocher, J, El-Sayed, M, Hazlett, HC & Styner, MA 2012, 'Asymmetric bias in user guided segmentations of brain structures', NeuroImage, vol. 59, no. 2, pp. 1315-1323. https://doi.org/10.1016/j.neuroimage.2011.08.025
Maltbie E, Bhatt K, Paniagua B, Smith RG, Graves MM, Mosconi MW et al. Asymmetric bias in user guided segmentations of brain structures. NeuroImage. 2012 Jan 16;59(2):1315-1323. https://doi.org/10.1016/j.neuroimage.2011.08.025
Maltbie, Eric ; Bhatt, Kshamta ; Paniagua, Beatriz ; Smith, Rachel G. ; Graves, Michael M. ; Mosconi, Matthew W. ; Peterson, Sarah ; White, Scott ; Blocher, Joseph ; El-Sayed, Mohammed ; Hazlett, Heather C. ; Styner, Martin A. / Asymmetric bias in user guided segmentations of brain structures. In: NeuroImage. 2012 ; Vol. 59, No. 2. pp. 1315-1323.
@article{a34c15b0069e4af7927b1a21dc0a987b,
title = "Asymmetric bias in user guided segmentations of brain structures",
abstract = "Brain morphometric studies often incorporate comparative hemispheric asymmetry analyses of segmented brain structures. In this work, we present evidence that common user guided structural segmentation techniques exhibit strong left-right asymmetric biases and thus fundamentally influence any left-right asymmetry analyses. In this study, MRI scans from ten pediatric subjects were employed for studying segmentations of amygdala, globus pallidus, putamen, caudate, and lateral ventricle. Additionally, two pediatric and three adult scans were used for studying hippocampus segmentation. Segmentations of the sub-cortical structures were performed by skilled raters using standard manual and semi-automated methods. The left-right mirrored versions of each image were included in the data and segmented in a random order to assess potential left-right asymmetric bias. Using shape analysis we further assessed whether the asymmetric bias is consistent across subjects and raters with the focus on the hippocampus. The user guided segmentation techniques on the sub-cortical structures exhibited left-right asymmetric volume bias with the hippocampus displaying the most significant asymmetry values (p << 0.01). The hippocampal shape analysis revealed the bias to be strongest on the lateral side of the body and medial side of the head and tail. The origin of this asymmetric bias is considered to be based in laterality of visual perception; therefore segmentations with any degree of user interaction contain an asymmetric bias. The aim of our study is to raise awareness in the neuroimaging community regarding the presence of the asymmetric bias and its influence on any left-right hemispheric analyses. We also recommend reexamining previous research results in the light of this new finding.",
keywords = "Asymmetry analysis, Manual segmentation, Rater bias, Subcortical structures neuroimaging",
author = "Eric Maltbie and Kshamta Bhatt and Beatriz Paniagua and Smith, {Rachel G.} and Graves, {Michael M.} and Mosconi, {Matthew W.} and Sarah Peterson and Scott White and Joseph Blocher and Mohammed El-Sayed and Hazlett, {Heather C.} and Styner, {Martin A.}",
year = "2012",
month = "1",
day = "16",
doi = "10.1016/j.neuroimage.2011.08.025",
language = "English (US)",
volume = "59",
pages = "1315--1323",
journal = "NeuroImage",
issn = "1053-8119",
publisher = "Academic Press Inc.",
number = "2",

}

TY - JOUR

T1 - Asymmetric bias in user guided segmentations of brain structures

AU - Maltbie, Eric

AU - Bhatt, Kshamta

AU - Paniagua, Beatriz

AU - Smith, Rachel G.

AU - Graves, Michael M.

AU - Mosconi, Matthew W.

AU - Peterson, Sarah

AU - White, Scott

AU - Blocher, Joseph

AU - El-Sayed, Mohammed

AU - Hazlett, Heather C.

AU - Styner, Martin A.

PY - 2012/1/16

Y1 - 2012/1/16

N2 - Brain morphometric studies often incorporate comparative hemispheric asymmetry analyses of segmented brain structures. In this work, we present evidence that common user guided structural segmentation techniques exhibit strong left-right asymmetric biases and thus fundamentally influence any left-right asymmetry analyses. In this study, MRI scans from ten pediatric subjects were employed for studying segmentations of amygdala, globus pallidus, putamen, caudate, and lateral ventricle. Additionally, two pediatric and three adult scans were used for studying hippocampus segmentation. Segmentations of the sub-cortical structures were performed by skilled raters using standard manual and semi-automated methods. The left-right mirrored versions of each image were included in the data and segmented in a random order to assess potential left-right asymmetric bias. Using shape analysis we further assessed whether the asymmetric bias is consistent across subjects and raters with the focus on the hippocampus. The user guided segmentation techniques on the sub-cortical structures exhibited left-right asymmetric volume bias with the hippocampus displaying the most significant asymmetry values (p << 0.01). The hippocampal shape analysis revealed the bias to be strongest on the lateral side of the body and medial side of the head and tail. The origin of this asymmetric bias is considered to be based in laterality of visual perception; therefore segmentations with any degree of user interaction contain an asymmetric bias. The aim of our study is to raise awareness in the neuroimaging community regarding the presence of the asymmetric bias and its influence on any left-right hemispheric analyses. We also recommend reexamining previous research results in the light of this new finding.

AB - Brain morphometric studies often incorporate comparative hemispheric asymmetry analyses of segmented brain structures. In this work, we present evidence that common user guided structural segmentation techniques exhibit strong left-right asymmetric biases and thus fundamentally influence any left-right asymmetry analyses. In this study, MRI scans from ten pediatric subjects were employed for studying segmentations of amygdala, globus pallidus, putamen, caudate, and lateral ventricle. Additionally, two pediatric and three adult scans were used for studying hippocampus segmentation. Segmentations of the sub-cortical structures were performed by skilled raters using standard manual and semi-automated methods. The left-right mirrored versions of each image were included in the data and segmented in a random order to assess potential left-right asymmetric bias. Using shape analysis we further assessed whether the asymmetric bias is consistent across subjects and raters with the focus on the hippocampus. The user guided segmentation techniques on the sub-cortical structures exhibited left-right asymmetric volume bias with the hippocampus displaying the most significant asymmetry values (p << 0.01). The hippocampal shape analysis revealed the bias to be strongest on the lateral side of the body and medial side of the head and tail. The origin of this asymmetric bias is considered to be based in laterality of visual perception; therefore segmentations with any degree of user interaction contain an asymmetric bias. The aim of our study is to raise awareness in the neuroimaging community regarding the presence of the asymmetric bias and its influence on any left-right hemispheric analyses. We also recommend reexamining previous research results in the light of this new finding.

KW - Asymmetry analysis

KW - Manual segmentation

KW - Rater bias

KW - Subcortical structures neuroimaging

UR - http://www.scopus.com/inward/record.url?scp=83055181250&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=83055181250&partnerID=8YFLogxK

U2 - 10.1016/j.neuroimage.2011.08.025

DO - 10.1016/j.neuroimage.2011.08.025

M3 - Article

VL - 59

SP - 1315

EP - 1323

JO - NeuroImage

JF - NeuroImage

SN - 1053-8119

IS - 2

ER -