Aarhus University Seal / Aarhus Universitets segl

Comparison of classification methods for tissue outcome after ischaemic stroke

Research output: Contribution to journal/Conference contribution in journal/Contribution to newspaperJournal articleResearchpeer-review

Standard

Comparison of classification methods for tissue outcome after ischaemic stroke. / Tozlu, Ceren; Ozenne, Brice; Cho, Tae Hee; Nighoghossian, Norbert; Mikkelsen, Irene Klærke; Derex, Laurent; Hermier, Marc; Pedraza, Salvador; Fiehler, Jens; Østergaard, Leif; Berthezène, Yves; Baron, Jean Claude; Maucort-Boulch, Delphine.

In: European Journal of Neuroscience, Vol. 50, No. 10, 11.2019, p. 3590-3598.

Research output: Contribution to journal/Conference contribution in journal/Contribution to newspaperJournal articleResearchpeer-review

Harvard

Tozlu, C, Ozenne, B, Cho, TH, Nighoghossian, N, Mikkelsen, IK, Derex, L, Hermier, M, Pedraza, S, Fiehler, J, Østergaard, L, Berthezène, Y, Baron, JC & Maucort-Boulch, D 2019, 'Comparison of classification methods for tissue outcome after ischaemic stroke', European Journal of Neuroscience, vol. 50, no. 10, pp. 3590-3598. https://doi.org/10.1111/ejn.14507

APA

Tozlu, C., Ozenne, B., Cho, T. H., Nighoghossian, N., Mikkelsen, I. K., Derex, L., Hermier, M., Pedraza, S., Fiehler, J., Østergaard, L., Berthezène, Y., Baron, J. C., & Maucort-Boulch, D. (2019). Comparison of classification methods for tissue outcome after ischaemic stroke. European Journal of Neuroscience, 50(10), 3590-3598. https://doi.org/10.1111/ejn.14507

CBE

Tozlu C, Ozenne B, Cho TH, Nighoghossian N, Mikkelsen IK, Derex L, Hermier M, Pedraza S, Fiehler J, Østergaard L, Berthezène Y, Baron JC, Maucort-Boulch D. 2019. Comparison of classification methods for tissue outcome after ischaemic stroke. European Journal of Neuroscience. 50(10):3590-3598. https://doi.org/10.1111/ejn.14507

MLA

Tozlu, Ceren et al. "Comparison of classification methods for tissue outcome after ischaemic stroke". European Journal of Neuroscience. 2019, 50(10). 3590-3598. https://doi.org/10.1111/ejn.14507

Vancouver

Tozlu C, Ozenne B, Cho TH, Nighoghossian N, Mikkelsen IK, Derex L et al. Comparison of classification methods for tissue outcome after ischaemic stroke. European Journal of Neuroscience. 2019 Nov;50(10):3590-3598. https://doi.org/10.1111/ejn.14507

Author

Tozlu, Ceren ; Ozenne, Brice ; Cho, Tae Hee ; Nighoghossian, Norbert ; Mikkelsen, Irene Klærke ; Derex, Laurent ; Hermier, Marc ; Pedraza, Salvador ; Fiehler, Jens ; Østergaard, Leif ; Berthezène, Yves ; Baron, Jean Claude ; Maucort-Boulch, Delphine. / Comparison of classification methods for tissue outcome after ischaemic stroke. In: European Journal of Neuroscience. 2019 ; Vol. 50, No. 10. pp. 3590-3598.

Bibtex

@article{6934cc55b04e485a86a903e0e0a5d169,
title = "Comparison of classification methods for tissue outcome after ischaemic stroke",
abstract = "In acute ischaemic stroke, identifying brain tissue at high risk of infarction is important for clinical decision-making. This tissue may be identified with suitable classification methods from magnetic resonance imaging data. The aim of the present study was to assess and compare the performance of five popular classification methods (adaptive boosting, logistic regression, artificial neural networks, random forest and support vector machine) in identifying tissue at high risk of infarction on human voxel-based brain imaging data. The classification methods were used with eight MRI parameters, including diffusion-weighted imaging and perfusion-weighted imaging obtained in 55 patients. The five criteria used to assess the performance of the methods were the area under the receiver operating curve (AUCroc), the area under the precision–recall curve (AUCpr), sensitivity, specificity and the Dice coefficient. The methods performed equally in terms of sensitivity and specificity, while the results of AUCroc and the Dice coefficient were significantly better for adaptive boosting, logistic regression, artificial neural networks and random forest. However, there was no statistically significant difference between the performances of these five classification methods regarding AUCpr, which was the main comparison metric. Machine learning methods can provide valuable prognostic information using multimodal imaging data in acute ischaemic stroke, which in turn can assist in developing personalized treatment decision for clinicians after a thorough validation of methods with an independent data set.",
keywords = "brain ischaemia, classification, diffusion-weighted imaging, machine learning, magnetic resonance imaging, perfusion-weighted imaging",
author = "Ceren Tozlu and Brice Ozenne and Cho, {Tae Hee} and Norbert Nighoghossian and Mikkelsen, {Irene Kl{\ae}rke} and Laurent Derex and Marc Hermier and Salvador Pedraza and Jens Fiehler and Leif {\O}stergaard and Yves Berthez{\`e}ne and Baron, {Jean Claude} and Delphine Maucort-Boulch",
year = "2019",
month = nov,
doi = "10.1111/ejn.14507",
language = "English",
volume = "50",
pages = "3590--3598",
journal = "European Journal of Neuroscience",
issn = "0953-816X",
publisher = "Wiley-Blackwell Publishing Ltd.",
number = "10",

}

RIS

TY - JOUR

T1 - Comparison of classification methods for tissue outcome after ischaemic stroke

AU - Tozlu, Ceren

AU - Ozenne, Brice

AU - Cho, Tae Hee

AU - Nighoghossian, Norbert

AU - Mikkelsen, Irene Klærke

AU - Derex, Laurent

AU - Hermier, Marc

AU - Pedraza, Salvador

AU - Fiehler, Jens

AU - Østergaard, Leif

AU - Berthezène, Yves

AU - Baron, Jean Claude

AU - Maucort-Boulch, Delphine

PY - 2019/11

Y1 - 2019/11

N2 - In acute ischaemic stroke, identifying brain tissue at high risk of infarction is important for clinical decision-making. This tissue may be identified with suitable classification methods from magnetic resonance imaging data. The aim of the present study was to assess and compare the performance of five popular classification methods (adaptive boosting, logistic regression, artificial neural networks, random forest and support vector machine) in identifying tissue at high risk of infarction on human voxel-based brain imaging data. The classification methods were used with eight MRI parameters, including diffusion-weighted imaging and perfusion-weighted imaging obtained in 55 patients. The five criteria used to assess the performance of the methods were the area under the receiver operating curve (AUCroc), the area under the precision–recall curve (AUCpr), sensitivity, specificity and the Dice coefficient. The methods performed equally in terms of sensitivity and specificity, while the results of AUCroc and the Dice coefficient were significantly better for adaptive boosting, logistic regression, artificial neural networks and random forest. However, there was no statistically significant difference between the performances of these five classification methods regarding AUCpr, which was the main comparison metric. Machine learning methods can provide valuable prognostic information using multimodal imaging data in acute ischaemic stroke, which in turn can assist in developing personalized treatment decision for clinicians after a thorough validation of methods with an independent data set.

AB - In acute ischaemic stroke, identifying brain tissue at high risk of infarction is important for clinical decision-making. This tissue may be identified with suitable classification methods from magnetic resonance imaging data. The aim of the present study was to assess and compare the performance of five popular classification methods (adaptive boosting, logistic regression, artificial neural networks, random forest and support vector machine) in identifying tissue at high risk of infarction on human voxel-based brain imaging data. The classification methods were used with eight MRI parameters, including diffusion-weighted imaging and perfusion-weighted imaging obtained in 55 patients. The five criteria used to assess the performance of the methods were the area under the receiver operating curve (AUCroc), the area under the precision–recall curve (AUCpr), sensitivity, specificity and the Dice coefficient. The methods performed equally in terms of sensitivity and specificity, while the results of AUCroc and the Dice coefficient were significantly better for adaptive boosting, logistic regression, artificial neural networks and random forest. However, there was no statistically significant difference between the performances of these five classification methods regarding AUCpr, which was the main comparison metric. Machine learning methods can provide valuable prognostic information using multimodal imaging data in acute ischaemic stroke, which in turn can assist in developing personalized treatment decision for clinicians after a thorough validation of methods with an independent data set.

KW - brain ischaemia

KW - classification

KW - diffusion-weighted imaging

KW - machine learning

KW - magnetic resonance imaging

KW - perfusion-weighted imaging

UR - http://www.scopus.com/inward/record.url?scp=85073781688&partnerID=8YFLogxK

U2 - 10.1111/ejn.14507

DO - 10.1111/ejn.14507

M3 - Journal article

C2 - 31278787

AN - SCOPUS:85073781688

VL - 50

SP - 3590

EP - 3598

JO - European Journal of Neuroscience

JF - European Journal of Neuroscience

SN - 0953-816X

IS - 10

ER -