Nonlinear smoothing of MR images using approximate entropy — A local measure of signal intensity irregularity

Geoffrey J.M. Parker, Julia A. Schnabel, Gareth J. Barker

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

4 Scopus citations

Abstract

Approximate entropy (ApEn) is a computable measure of sequential irregularity that is applicable to sequences of numbers of finite length. As such, it may be used to determine how random a sequence of numbers is. We exploit this property to determine the relevance of image information; to determine whether a spatial signal intensity distribution varies in a regular fashion — and is therefore likely to be an image feature or image texture, or is highly random — and likely to be noise. We present an outline of two possible methodologies for creating an ApEn-based noise filter: a modified median filter and a modified anisotropic diffusion scheme. We show that both approaches lead to effective noise reduction in MR images, with improved information-retaining properties when compared with their conventional counterparts.

Original languageEnglish
Title of host publicationInformation Processing in Medical Imaging - 16th InternationalConference, IPMI 1999, Proceedings
EditorsAttila Kuba, Martin Samal, Andrew Todd-Pokropek
PublisherSpringer Verlag
Pages484-489
Number of pages6
ISBN (Print)3540661670, 9783540661672
DOIs
StatePublished - 1999
Externally publishedYes
Event16th International conference on Information Processing in Medical Imaging, IPMI 1999 - Visegrad, Hungary
Duration: 28 Jun 19992 Jul 1999

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume1613
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference16th International conference on Information Processing in Medical Imaging, IPMI 1999
Country/TerritoryHungary
CityVisegrad
Period28/06/992/07/99

Fingerprint

Dive into the research topics of 'Nonlinear smoothing of MR images using approximate entropy — A local measure of signal intensity irregularity'. Together they form a unique fingerprint.

Cite this