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

Ona Wu; Stefan Winzeck; Anne Katrin Giese; Brandon L. Hancock; Mark R. Etherton; Mark J.R.J. Bouts; Kathleen Donahue; Markus D. Schirmer; Robert E. Irie; Steven J.T. Mocking; Elissa C. McIntosh; Raquel Bezerra; Konstantinos Kamnitsas; Petrea Frid; Johan Wasselius; John W. Cole; Huichun Xu; Lukas Holmegaard; Jordi Jiménez-Conde; Robin Lemmens; Eric Lorentzen; Patrick F. McArdle; James F. Meschia; Jaume Roquer; Tatjana Rundek; Ralph L. Sacco; Reinhold Schmidt; Pankaj Sharma; Agnieszka Slowik; Tara M. Stanne; Vincent Thijs; Achala Vagal; Daniel Woo; Stephen Bevan; Steven J. Kittner; Braxton D. Mitchell; Jonathan Rosand; Bradford B. Worrall; Christina Jern; Arne G. Lindgren; Jane Maguire; Natalia S. Rost

doi:10.1161/STROKEAHA.119.025373

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

Ona Wu, Stefan Winzeck, Anne Katrin Giese, Brandon L. Hancock, Mark R. Etherton, Mark J.R.J. Bouts, Kathleen Donahue, Markus D. Schirmer, Robert E. Irie, Steven J.T. Mocking, Elissa C. McIntosh, Raquel Bezerra, Konstantinos Kamnitsas, Petrea Frid, Johan Wasselius, John W. Cole, Huichun Xu, Lukas Holmegaard, Jordi Jiménez-Conde, Robin LemmensEric Lorentzen, Patrick F. McArdle, James F. Meschia, Jaume Roquer, Tatjana Rundek, Ralph L. Sacco, Reinhold Schmidt, Pankaj Sharma, Agnieszka Slowik, Tara M. Stanne, Vincent Thijs, Achala Vagal, Daniel Woo, Stephen Bevan, Steven J. Kittner, Braxton D. Mitchell, Jonathan Rosand, Bradford B. Worrall, Christina Jern, Arne G. Lindgren, Jane Maguire, Natalia S. Rost

Neurology

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

10 Scopus citations

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 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.

Original language	English (US)
Pages (from-to)	1734-1741
Number of pages	8
Journal	Stroke
Volume	50
Issue number	7
DOIs	https://doi.org/10.1161/STROKEAHA.119.025373
State	Published - Jul 1 2019

Keywords

diffusion magnetic resonance imaging
machine learning
phenotype
risk factors
stroke

ASJC Scopus subject areas

Clinical Neurology
Cardiology and Cardiovascular Medicine
Advanced and Specialized Nursing

Access to Document

10.1161/STROKEAHA.119.025373

Cite this

Wu, O., Winzeck, S., Giese, A. K., Hancock, B. L., Etherton, M. R., Bouts, M. J. R. J., Donahue, K., Schirmer, M. D., Irie, R. E., Mocking, S. J. T., McIntosh, E. C., Bezerra, R., Kamnitsas, K., Frid, P., Wasselius, J., Cole, J. W., Xu, H., Holmegaard, L., Jiménez-Conde, J., ... Rost, N. S. (2019). Big Data Approaches to Phenotyping Acute Ischemic Stroke Using Automated Lesion Segmentation of Multi-Center Magnetic Resonance Imaging Data. Stroke, 50(7), 1734-1741. https://doi.org/10.1161/STROKEAHA.119.025373

Wu, O, Winzeck, S, Giese, AK, Hancock, BL, Etherton, MR, Bouts, MJRJ, Donahue, K, Schirmer, MD, Irie, RE, Mocking, SJT, McIntosh, EC, Bezerra, R, Kamnitsas, K, Frid, P, Wasselius, J, Cole, JW, Xu, H, Holmegaard, L, Jiménez-Conde, J, Lemmens, R, Lorentzen, E, McArdle, PF, Meschia, JF, Roquer, J, Rundek, T, Sacco, RL, Schmidt, R, Sharma, P, Slowik, A, Stanne, TM, Thijs, V, Vagal, A, Woo, D, Bevan, S, Kittner, SJ, Mitchell, BD, Rosand, J, Worrall, BB, Jern, C, Lindgren, AG, Maguire, J & Rost, NS 2019, 'Big Data Approaches to Phenotyping Acute Ischemic Stroke Using Automated Lesion Segmentation of Multi-Center Magnetic Resonance Imaging Data', Stroke, vol. 50, no. 7, pp. 1734-1741. https://doi.org/10.1161/STROKEAHA.119.025373

@article{f29de9b48b7f40cd9aa145fe82638749,

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

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 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.",

keywords = "diffusion magnetic resonance imaging, machine learning, phenotype, risk factors, stroke",

author = "Ona Wu and Stefan Winzeck and Giese, {Anne Katrin} and Hancock, {Brandon L.} and Etherton, {Mark R.} and Bouts, {Mark J.R.J.} and Kathleen Donahue and Schirmer, {Markus D.} and Irie, {Robert E.} and Mocking, {Steven J.T.} and McIntosh, {Elissa C.} and Raquel Bezerra and Konstantinos Kamnitsas and Petrea Frid and Johan Wasselius and Cole, {John W.} and Huichun Xu and Lukas Holmegaard and Jordi Jim{\'e}nez-Conde and Robin Lemmens and Eric Lorentzen and McArdle, {Patrick F.} and Meschia, {James F.} and Jaume Roquer and Tatjana Rundek and Sacco, {Ralph L.} and Reinhold Schmidt and Pankaj Sharma and Agnieszka Slowik and Stanne, {Tara M.} and Vincent Thijs and Achala Vagal and Daniel Woo and Stephen Bevan and Kittner, {Steven J.} and Mitchell, {Braxton D.} and Jonathan Rosand and Worrall, {Bradford B.} and Christina Jern and Lindgren, {Arne G.} and Jane Maguire and Rost, {Natalia S.}",

note = "Publisher Copyright: {\textcopyright} 2019 American Heart Association, Inc.",

year = "2019",

month = jul,

day = "1",

doi = "10.1161/STROKEAHA.119.025373",

language = "English (US)",

volume = "50",

pages = "1734--1741",

journal = "Stroke",

issn = "0039-2499",

publisher = "Lippincott Williams and Wilkins",

number = "7",

}

TY - JOUR

T1 - Big Data Approaches to Phenotyping Acute Ischemic Stroke Using Automated Lesion Segmentation of Multi-Center Magnetic Resonance Imaging Data

AU - Wu, Ona

AU - Winzeck, Stefan

AU - Giese, Anne Katrin

AU - Hancock, Brandon L.

AU - Etherton, Mark R.

AU - Bouts, Mark J.R.J.

AU - Donahue, Kathleen

AU - Schirmer, Markus D.

AU - Irie, Robert E.

AU - Mocking, Steven J.T.

AU - McIntosh, Elissa C.

AU - Bezerra, Raquel

AU - Kamnitsas, Konstantinos

AU - Frid, Petrea

AU - Wasselius, Johan

AU - Cole, John W.

AU - Xu, Huichun

AU - Holmegaard, Lukas

AU - Jiménez-Conde, Jordi

AU - Lemmens, Robin

AU - Lorentzen, Eric

AU - McArdle, Patrick F.

AU - Meschia, James F.

AU - Roquer, Jaume

AU - Rundek, Tatjana

AU - Sacco, Ralph L.

AU - Schmidt, Reinhold

AU - Sharma, Pankaj

AU - Slowik, Agnieszka

AU - Stanne, Tara M.

AU - Thijs, Vincent

AU - Vagal, Achala

AU - Woo, Daniel

AU - Bevan, Stephen

AU - Kittner, Steven J.

AU - Mitchell, Braxton D.

AU - Rosand, Jonathan

AU - Worrall, Bradford B.

AU - Jern, Christina

AU - Lindgren, Arne G.

AU - Maguire, Jane

AU - Rost, Natalia S.

PY - 2019/7/1

Y1 - 2019/7/1

N2 - 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.

AB - 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.

KW - diffusion magnetic resonance imaging

KW - machine learning

KW - phenotype

KW - risk factors

KW - stroke

UR - http://www.scopus.com/inward/record.url?scp=85068808645&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85068808645&partnerID=8YFLogxK

U2 - 10.1161/STROKEAHA.119.025373

DO - 10.1161/STROKEAHA.119.025373

M3 - Article

C2 - 31177973

AN - SCOPUS:85068808645

SN - 0039-2499

VL - 50

SP - 1734

EP - 1741

JO - Stroke

JF - Stroke

IS - 7

ER -

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

Abstract

Keywords

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this