Fulltext

Comparative Performance Analysis of State-of-the-Art Classification Algorithms Applied to Lung Tissue Categorization

Depeursinge, Adrien ; Iavindrasana, Jimison ; Hidki, Asmâa ; Cohen, Gilles ; Geissbuhler, Antoine ; Platon, Alexandra ; Poletti, Pierre-Alexandre ; Müller, Henning

In: Journal of Digital Imaging, 2010, vol. 23, no. 1, p. 18-30

Ajouter à la liste personnelle
    Titre
    • Comparative Performance Analysis of State-of-the-Art Classification Algorithms Applied to Lung Tissue Categorization
    Auteur
    • Depeursinge, Adrien. Service of Medical Informatics, Geneva University Hospitals and University of Geneva, 24, rue Micheli-du-Crest, CH-1211, Geneva 14, Switzerland
    • Iavindrasana, Jimison. Service of Medical Informatics, Geneva University Hospitals and University of Geneva, 24, rue Micheli-du-Crest, CH-1211, Geneva 14, Switzerland
    • Hidki, Asmâa. Service of Medical Informatics, Geneva University Hospitals and University of Geneva, 24, rue Micheli-du-Crest, CH-1211, Geneva 14, Switzerland
    • Cohen, Gilles. Service of Medical Informatics, Geneva University Hospitals and University of Geneva, 24, rue Micheli-du-Crest, CH-1211, Geneva 14, Switzerland
    • Geissbuhler, Antoine. Service of Medical Informatics, Geneva University Hospitals and University of Geneva, 24, rue Micheli-du-Crest, CH-1211, Geneva 14, Switzerland
    • Platon, Alexandra. Service of Emergency Radiology, Geneva University Hospitals and University of Geneva, 24, rue Micheli-du-Crest, CH-1211, Geneva 14, Switzerland
    • Poletti, Pierre-Alexandre. Service of Emergency Radiology, Geneva University Hospitals and University of Geneva, 24, rue Micheli-du-Crest, CH-1211, Geneva 14, Switzerland
    • Müller, Henning. Service of Medical Informatics, Geneva University Hospitals and University of Geneva, 24, rue Micheli-du-Crest, CH-1211, Geneva 14, Switzerland
    Type de document
    • Postprint
    Identifiant OAI-PMH
    • oai:doc.rero.ch:313970
    Summary
    In this paper, we compare five common classifier families in their ability to categorize six lung tissue patterns in high-resolution computed tomography (HRCT) images of patients affected with interstitial lung diseases (ILD) and with healthy tissue. The evaluated classifiers are naive Bayes, k-nearest neighbor, J48 decision trees, multilayer perceptron, and support vector machines (SVM). The dataset used contains 843 regions of interest (ROI) of healthy and five pathologic lung tissue patterns identified by two radiologists at the University Hospitals of Geneva. Correlation of the feature space composed of 39 texture attributes is studied. A grid search for optimal parameters is carried out for each classifier family. Two complementary metrics are used to characterize the performances of classification. These are based on McNemar's statistical tests and global accuracy. SVM reached best values for each metric and allowed a mean correct prediction rate of 88.3% with high class-specific precision on testing sets of 423 ROIs