Wavelets and fuzzy relational classifiers: A novel spectroscopy analysis system for pediatric metabolic brain diseases

  • Authors:
  • S. Zarei Mahmoodabadi;J. Alirezaie;P. Babyn;A. Kassner;E. Widjaja

  • Affiliations:
  • Department of Electrical and Computer Engineering, Ryerson University, Toronto, Canada and Department of Diagnostic Imaging, The Hospital for Sick Children, Toronto, Canada;Department of Electrical and Computer Engineering, Ryerson University, Toronto, Canada;Department of Diagnostic Imaging, The Hospital for Sick Children, Toronto, Canada;Department of Medical Imaging, University of Toronto, Toronto, Canada;Department of Diagnostic Imaging, The Hospital for Sick Children, Toronto, Canada

  • Venue:
  • Fuzzy Sets and Systems
  • Year:
  • 2010

Quantified Score

Hi-index 0.20

Visualization

Abstract

A suspected metabolic disorder presents a difficult challenge to the physician and the patient. We have developed a fully automated system in order to analyze and classify the magnetic resonance spectroscopy signals of patients with metabolic brain diseases. We utilized wavelets to extract signal features and in the time-frequency representations to optimize the feature extraction procedure. Novel fuzzy membership functions and a fuzzy relational classifier were designed to categorize the metabolic brain diseases in children using the information obtained from the feature extraction routine. The sensitivity (Se) and the positive predictivity (PP) of 88.26% and 91.04% in extracting features and 89.66% and 100%, respectively, in detecting metabolic brain diseases has been achieved.