Skip to main navigation Skip to search Skip to main content

Texture-based continuous probabilistic framework for robust medical image representation and classification

  • Dror Lederman

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

This paper presents a texture-based continuous probabilistic framework for robust image representation. According to the proposed approach, images taken at different angles are represented using several probabilistic models connected in parallel. The classification decision is made based on a maximum likelihood approach, which is insensitive to the angle at which the image was taken. The proposed approach is evaluated using a dataset of 100 images that includes three classes of anatomical structures of the upper airways. The results show that the approach can be used to efficiently and reliably represent and classify medical images acquired during various procedures.

Original languageEnglish
Title of host publication6th European Conference of the International Federation for Medical and Biological Engineering - MBEC 2014
EditorsIgor Lackovic, Darko Vasic
PublisherSpringer Verlag
Pages196-199
Number of pages4
ISBN (Electronic)9783319111278
DOIs
StatePublished - 1 Jan 2015
Externally publishedYes
Event6th European Conference of the International Federation for Medical and Biological Engineering, MBEC 2014 - Dubrovnik, Croatia
Duration: 7 Sep 201411 Sep 2014

Publication series

NameIFMBE Proceedings
Volume45
ISSN (Print)1680-0737

Conference

Conference6th European Conference of the International Federation for Medical and Biological Engineering, MBEC 2014
Country/TerritoryCroatia
CityDubrovnik
Period7/09/1411/09/14

Keywords

  • Classification
  • Gaussian mixture models
  • Medical imaging
  • Textural features

ASJC Scopus subject areas

  • Bioengineering
  • Biomedical Engineering

Fingerprint

Dive into the research topics of 'Texture-based continuous probabilistic framework for robust medical image representation and classification'. Together they form a unique fingerprint.

Cite this