Analysis of intensity normalization for optimal segmentation performance of a fully convolutional neural network.

Journal: Zeitschrift fur medizinische Physik
Published Date:

Abstract

INTRODUCTION: Convolutional neural networks have begun to surpass classical statistical- and atlas based machine learning techniques in medical image segmentation in recent years, proving to be superior in performance and speed. However, a major challenge that the community faces are mismatch between variability within training and evaluation datasets and therefore a dependency on proper data pre-processing. Intensity normalization is a widely applied technique for reducing the variance of the data for which there are several methods available ranging from uniformity transformation to histogram equalization. The current study analyses the influence of intensity normalization on cerebellum segmentation performance of a convolutional neural network (CNN).

Authors

  • Nina Jacobsen
    Medical Physics Group, Institute for Diagnostic and Interventional Radiology, University Hospital Jena, Jena, Germany.
  • Andreas Deistung
    Medical Physics Group, Institute for Diagnostic and Interventional Radiology, University Hospital Jena, Jena, Germany; Department of Neurology, Essen University Hospital, University of Duisburg-Essen, Essen, Germany.
  • Dagmar Timmann
    Department of Neurology, Essen University Hospital, University of Duisburg-Essen, Essen, Germany.
  • Sophia L Goericke
    Department of Diagnostic and Interventional Radiology and Neuroradiology, University of Duisburg-Essen, Essen, Germany.
  • Jürgen R Reichenbach
    Medical Physics Group, Institute for Diagnostic and Interventional Radiology, University Hospital Jena, Jena, Germany; Michael Stifel Center for Data-Driven and Simulation Science, Friedrich Schiller University Jena, Jena, Germany.
  • Daniel Güllmar
    Medical Physics Group, Institute for Diagnostic and Interventional Radiology, University Hospital Jena, Jena, Germany. Electronic address: daniel.guellmar@med.uni-jena.de.