Skip to main content

Lund University Publications

LUND UNIVERSITY LIBRARIES

Big Data Approaches to Phenotyping Acute Ischemic Stroke Using Automated Lesion Segmentation of Multi-Center Magnetic Resonance Imaging Data

Wu, Ona ; Winzeck, Stefan ; Giese, Anne-Katrin ; Hancock, Brandon L ; Etherton, Mark R ; Bouts, Mark J R J ; Donahue, Kathleen ; Schirmer, Markus D ; Irie, Robert E and Mocking, Steven J T , et al. (2019) In Stroke 50(7). p.1734-1741
Abstract

Background and Purpose- We evaluated deep learning algorithms' segmentation of acute ischemic lesions on heterogeneous multi-center clinical diffusion-weighted magnetic resonance imaging (MRI) data sets and explored the potential role of this tool for phenotyping acute ischemic stroke. Methods- Ischemic stroke data sets from the MRI-GENIE (MRI-Genetics Interface Exploration) repository consisting of 12 international genetic research centers were retrospectively analyzed using an automated deep learning segmentation algorithm consisting of an ensemble of 3-dimensional convolutional neural networks. Three ensembles were trained using data from the following: (1) 267 patients from an independent single-center cohort, (2) 267 patients from... (More)

Background and Purpose- We evaluated deep learning algorithms' segmentation of acute ischemic lesions on heterogeneous multi-center clinical diffusion-weighted magnetic resonance imaging (MRI) data sets and explored the potential role of this tool for phenotyping acute ischemic stroke. Methods- Ischemic stroke data sets from the MRI-GENIE (MRI-Genetics Interface Exploration) repository consisting of 12 international genetic research centers were retrospectively analyzed using an automated deep learning segmentation algorithm consisting of an ensemble of 3-dimensional convolutional neural networks. Three ensembles were trained using data from the following: (1) 267 patients from an independent single-center cohort, (2) 267 patients from MRI-GENIE, and (3) mixture of (1) and (2). The algorithms' performances were compared against manual outlines from a separate 383 patient subset from MRI-GENIE. Univariable and multivariable logistic regression with respect to demographics, stroke subtypes, and vascular risk factors were performed to identify phenotypes associated with large acute diffusion-weighted MRI volumes and greater stroke severity in 2770 MRI-GENIE patients. Stroke topography was investigated. Results- The ensemble consisting of a mixture of MRI-GENIE and single-center convolutional neural networks performed best. Subset analysis comparing automated and manual lesion volumes in 383 patients found excellent correlation (ρ=0.92; P<0.0001). Median (interquartile range) diffusion-weighted MRI lesion volumes from 2770 patients were 3.7 cm3 (0.9-16.6 cm3). Patients with small artery occlusion stroke subtype had smaller lesion volumes ( P<0.0001) and different topography compared with other stroke subtypes. Conclusions- Automated accurate clinical diffusion-weighted MRI lesion segmentation using deep learning algorithms trained with multi-center and diverse data is feasible. Both lesion volume and topography can provide insight into stroke subtypes with sufficient sample size from big heterogeneous multi-center clinical imaging phenotype data sets.

(Less)
Please use this url to cite or link to this publication:
author
; ; ; ; ; ; ; ; and , et al. (More)
; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; ; and (Less)
author collaboration
organization
publishing date
type
Contribution to journal
publication status
published
subject
in
Stroke
volume
50
issue
7
pages
1734 - 1741
publisher
American Heart Association
external identifiers
  • pmid:31177973
  • scopus:85068808645
ISSN
1524-4628
DOI
10.1161/STROKEAHA.119.025373
language
English
LU publication?
yes
id
4fd58602-8532-4498-b8b2-7070f9d21d2f
date added to LUP
2019-06-14 14:13:43
date last changed
2024-03-19 12:02:07
@article{4fd58602-8532-4498-b8b2-7070f9d21d2f,
  abstract     = {{<p>Background and Purpose- We evaluated deep learning algorithms' segmentation of acute ischemic lesions on heterogeneous multi-center clinical diffusion-weighted magnetic resonance imaging (MRI) data sets and explored the potential role of this tool for phenotyping acute ischemic stroke. Methods- Ischemic stroke data sets from the MRI-GENIE (MRI-Genetics Interface Exploration) repository consisting of 12 international genetic research centers were retrospectively analyzed using an automated deep learning segmentation algorithm consisting of an ensemble of 3-dimensional convolutional neural networks. Three ensembles were trained using data from the following: (1) 267 patients from an independent single-center cohort, (2) 267 patients from MRI-GENIE, and (3) mixture of (1) and (2). The algorithms' performances were compared against manual outlines from a separate 383 patient subset from MRI-GENIE. Univariable and multivariable logistic regression with respect to demographics, stroke subtypes, and vascular risk factors were performed to identify phenotypes associated with large acute diffusion-weighted MRI volumes and greater stroke severity in 2770 MRI-GENIE patients. Stroke topography was investigated. Results- The ensemble consisting of a mixture of MRI-GENIE and single-center convolutional neural networks performed best. Subset analysis comparing automated and manual lesion volumes in 383 patients found excellent correlation (ρ=0.92; P&lt;0.0001). Median (interquartile range) diffusion-weighted MRI lesion volumes from 2770 patients were 3.7 cm3 (0.9-16.6 cm3). Patients with small artery occlusion stroke subtype had smaller lesion volumes ( P&lt;0.0001) and different topography compared with other stroke subtypes. Conclusions- Automated accurate clinical diffusion-weighted MRI lesion segmentation using deep learning algorithms trained with multi-center and diverse data is feasible. Both lesion volume and topography can provide insight into stroke subtypes with sufficient sample size from big heterogeneous multi-center clinical imaging phenotype data sets.</p>}},
  author       = {{Wu, Ona and Winzeck, Stefan and Giese, Anne-Katrin and Hancock, Brandon L and Etherton, Mark R and Bouts, Mark J R J and Donahue, Kathleen and Schirmer, Markus D and Irie, Robert E and Mocking, Steven J T and McIntosh, Elissa C and Bezerra, Raquel and Kamnitsas, Konstantinos and Frid, Petrea and Wasselius, Johan and Cole, John W and Xu, Huichun and Holmegaard, Lukas and Jiménez-Conde, Jordi and Lemmens, Robin and Lorentzen, Eric and McArdle, Patrick F and Meschia, James F and Roquer, Jaume and Rundek, Tatjana and Sacco, Ralph L and Schmidt, Reinhold and Sharma, Pankaj and Slowik, Agnieszka and Stanne, Tara M and Thijs, Vincent and Vagal, Achala and Woo, Daniel and Bevan, Stephen and Kittner, Steven J and Mitchell, Braxton D and Rosand, Jonathan and Worrall, Bradford B and Jern, Christina and Lindgren, Arne G and Maguire, Jane and Rost, Natalia S}},
  issn         = {{1524-4628}},
  language     = {{eng}},
  month        = {{06}},
  number       = {{7}},
  pages        = {{1734--1741}},
  publisher    = {{American Heart Association}},
  series       = {{Stroke}},
  title        = {{Big Data Approaches to Phenotyping Acute Ischemic Stroke Using Automated Lesion Segmentation of Multi-Center Magnetic Resonance Imaging Data}},
  url          = {{http://dx.doi.org/10.1161/STROKEAHA.119.025373}},
  doi          = {{10.1161/STROKEAHA.119.025373}},
  volume       = {{50}},
  year         = {{2019}},
}