Forskning
Udskriv Udskriv
Switch language
Region Hovedstaden - en del af Københavns Universitetshospital
Udgivet

Comparison of classification methods for tissue outcome after ischaemic stroke

Publikation: Bidrag til tidsskriftTidsskriftartikelForskningpeer review

DOI

  1. Cortical oscillations and entrainment in speech processing during working memory load

    Publikation: Bidrag til tidsskriftTidsskriftartikelForskningpeer review

  2. Increasing propensity to mind-wander by transcranial direct current stimulation? A registered report

    Publikation: Bidrag til tidsskriftTidsskriftartikelForskningpeer review

  3. Blinding is compromised for transcranial direct current stimulation at 1 mA for 20 minutes in young healthy adults

    Publikation: Bidrag til tidsskriftTidsskriftartikelForskningpeer review

  4. Individualized quantification of the benefit from reperfusion therapy using stroke predictive models

    Publikation: Bidrag til tidsskriftTidsskriftartikelForskningpeer review

  5. Spatiotemporal expression pattern of PERIOD 1 and PERIOD 2 in the mouse SCN is dependent on VIP receptor 2 signaling

    Publikation: Bidrag til tidsskriftTidsskriftartikelForskningpeer review

  1. Assessing absolute stroke risk in patients with atrial fibrillation using a risk factor based approach

    Publikation: Bidrag til tidsskriftTidsskriftartikelForskningpeer review

  2. Brain serotonin 2A receptor binding predicts subjective temporal and mystical effects of psilocybin in healthy humans

    Publikation: Bidrag til tidsskriftTidsskriftartikelForskningpeer review

  3. Reward processing in major depressive disorder and prediction of treatment response - Neuropharm study

    Publikation: Bidrag til tidsskriftTidsskriftartikelForskningpeer review

  4. Net benefit in the presence of correlated prioritized outcomes using generalized pairwise comparisons: A simulation study

    Publikation: Bidrag til tidsskriftTidsskriftartikelForskningpeer review

  5. A new measure of treatment effect in clinical trials involving competing risks based on generalized pairwise comparisons

    Publikation: Bidrag til tidsskriftTidsskriftartikelForskningpeer review

  • Ceren Tozlu
  • Brice Ozenne
  • Tae-Hee Cho
  • Norbert Nighoghossian
  • Irene Klaerke Mikkelsen
  • Laurent Derex
  • Marc Hermier
  • Salvador Pedraza
  • Jens Fiehler
  • Leif Østergaard
  • Yves Berthezène
  • Jean-Claude Baron
  • Delphine Maucort-Boulch
Vis graf over relationer

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.

OriginalsprogEngelsk
TidsskriftEuropean Journal of Neuroscience
Vol/bind50
Udgave nummer10
Sider (fra-til)3590-3598
Antal sider9
ISSN0953-816X
DOI
StatusUdgivet - nov. 2019

Bibliografisk note

© 2019 Federation of European Neuroscience Societies and John Wiley & Sons Ltd.

ID: 59018136