Validation and comparison of two automated methods to quantify brain white matter hyperintensities of presumed vascular origin

Jennifer M. J. Waymont (Corresponding Author), Chariklia Petsa, Chris J. McNeil, Alison D. Murray, Gordon D. Waiter

Research output: Contribution to journalArticle

Abstract

Background: White matter hyperintensities (WMH) are a common imaging finding, indicative of cerebral small vessel disease. Lesion segmentation algorithms have been developed to overcome issues arising from visual rating scales. In this study, two automated methods were evaluated against visual and manual segmentation. We aimed to determine the most robust algorithm provided by the open-source Lesion Segmentation Toolbox (LST). Methods: We compared WMH data from visual rating (Scheltens’ scale) with those derived from algorithms provided within LST. We then compared spatial and volumetric WMH data derived from manually-delineated lesion maps with WMH data and lesion maps provided by LST algorithms. Results: We identified optimal initial thresholds for algorithms provided by LST when compared with visual rating (LGA: initial K and lesion probability thresholds = 0.5; LPA lesion probability threshold = 0.65). LGA was found to perform best in comparisons with manual segmentation. Discussion: When compared to Scheltens’ score and to manual segmentation, LGA appeared to be the most suitable algorithm. We found LGA to be a userfriendly, effective WMH segmentation method for a research environment.
Original languageEnglish
JournalJournal of International Medical Research
Publication statusAccepted/In press - 12 Sep 2019

Fingerprint

Blood Vessels
Brain
Cerebral Small Vessel Diseases
White Matter
Research

Keywords

  • white matter hyperintensity
  • lesion segmentation
  • cerebral small vessel disease
  • brain ageing
  • methodology
  • validation

Cite this

@article{81ebcb0e5dfa411580faba0e2302027a,
title = "Validation and comparison of two automated methods to quantify brain white matter hyperintensities of presumed vascular origin",
abstract = "Background: White matter hyperintensities (WMH) are a common imaging finding, indicative of cerebral small vessel disease. Lesion segmentation algorithms have been developed to overcome issues arising from visual rating scales. In this study, two automated methods were evaluated against visual and manual segmentation. We aimed to determine the most robust algorithm provided by the open-source Lesion Segmentation Toolbox (LST). Methods: We compared WMH data from visual rating (Scheltens’ scale) with those derived from algorithms provided within LST. We then compared spatial and volumetric WMH data derived from manually-delineated lesion maps with WMH data and lesion maps provided by LST algorithms. Results: We identified optimal initial thresholds for algorithms provided by LST when compared with visual rating (LGA: initial K and lesion probability thresholds = 0.5; LPA lesion probability threshold = 0.65). LGA was found to perform best in comparisons with manual segmentation. Discussion: When compared to Scheltens’ score and to manual segmentation, LGA appeared to be the most suitable algorithm. We found LGA to be a userfriendly, effective WMH segmentation method for a research environment.",
keywords = "white matter hyperintensity, lesion segmentation, cerebral small vessel disease, brain ageing, methodology, validation",
author = "Waymont, {Jennifer M. J.} and Chariklia Petsa and McNeil, {Chris J.} and Murray, {Alison D.} and Waiter, {Gordon D.}",
note = "Funding Data collection was funded by grants from the Alzheimer’s Research Trust (now Alzheimer’s Research UK, grant reference: ART/SPG2003B), Alzheimer’s Research UK (grant reference: ARUK-SB2012B-2), the University of Aberdeen Development Trust (grant reference RGB3109) and NHS Grampian and the Chief Scientist’s Office (grant reference: CAF/08/08). JMJW is funded by the University of Aberdeen Development Trust and TauRx Therapeutics Ltd. CP is funded by Royal Surrey County Hospital NHS Foundation Trust. CJM, ADM, and GDW are funded by the Scottish Funding Council.",
year = "2019",
month = "9",
day = "12",
language = "English",
journal = "Journal of International Medical Research",
issn = "0300-0605",
publisher = "Sage Publications",

}

TY - JOUR

T1 - Validation and comparison of two automated methods to quantify brain white matter hyperintensities of presumed vascular origin

AU - Waymont, Jennifer M. J.

AU - Petsa, Chariklia

AU - McNeil, Chris J.

AU - Murray, Alison D.

AU - Waiter, Gordon D.

N1 - Funding Data collection was funded by grants from the Alzheimer’s Research Trust (now Alzheimer’s Research UK, grant reference: ART/SPG2003B), Alzheimer’s Research UK (grant reference: ARUK-SB2012B-2), the University of Aberdeen Development Trust (grant reference RGB3109) and NHS Grampian and the Chief Scientist’s Office (grant reference: CAF/08/08). JMJW is funded by the University of Aberdeen Development Trust and TauRx Therapeutics Ltd. CP is funded by Royal Surrey County Hospital NHS Foundation Trust. CJM, ADM, and GDW are funded by the Scottish Funding Council.

PY - 2019/9/12

Y1 - 2019/9/12

N2 - Background: White matter hyperintensities (WMH) are a common imaging finding, indicative of cerebral small vessel disease. Lesion segmentation algorithms have been developed to overcome issues arising from visual rating scales. In this study, two automated methods were evaluated against visual and manual segmentation. We aimed to determine the most robust algorithm provided by the open-source Lesion Segmentation Toolbox (LST). Methods: We compared WMH data from visual rating (Scheltens’ scale) with those derived from algorithms provided within LST. We then compared spatial and volumetric WMH data derived from manually-delineated lesion maps with WMH data and lesion maps provided by LST algorithms. Results: We identified optimal initial thresholds for algorithms provided by LST when compared with visual rating (LGA: initial K and lesion probability thresholds = 0.5; LPA lesion probability threshold = 0.65). LGA was found to perform best in comparisons with manual segmentation. Discussion: When compared to Scheltens’ score and to manual segmentation, LGA appeared to be the most suitable algorithm. We found LGA to be a userfriendly, effective WMH segmentation method for a research environment.

AB - Background: White matter hyperintensities (WMH) are a common imaging finding, indicative of cerebral small vessel disease. Lesion segmentation algorithms have been developed to overcome issues arising from visual rating scales. In this study, two automated methods were evaluated against visual and manual segmentation. We aimed to determine the most robust algorithm provided by the open-source Lesion Segmentation Toolbox (LST). Methods: We compared WMH data from visual rating (Scheltens’ scale) with those derived from algorithms provided within LST. We then compared spatial and volumetric WMH data derived from manually-delineated lesion maps with WMH data and lesion maps provided by LST algorithms. Results: We identified optimal initial thresholds for algorithms provided by LST when compared with visual rating (LGA: initial K and lesion probability thresholds = 0.5; LPA lesion probability threshold = 0.65). LGA was found to perform best in comparisons with manual segmentation. Discussion: When compared to Scheltens’ score and to manual segmentation, LGA appeared to be the most suitable algorithm. We found LGA to be a userfriendly, effective WMH segmentation method for a research environment.

KW - white matter hyperintensity

KW - lesion segmentation

KW - cerebral small vessel disease

KW - brain ageing

KW - methodology

KW - validation

M3 - Article

JO - Journal of International Medical Research

JF - Journal of International Medical Research

SN - 0300-0605

ER -