Multilayer perceptron based level sets for robust ultrasound image segmentation

M. Mora*, C. Tauber, H. Batatia

*Corresponding author for this work

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

The class of geometric deformable models, also known as level sets, has brought tremendous impact on medical imagery due to its capability of topology preservation and fast shape recovery. Ultrasound images are often characterized by a high level of speckle causing erroneous detection of contours. This work proposes a new stopping term for level sets, based on the coefficient of variation and a multilayer perceptron, in order to robustly detect the contours in ultrasound images. Successful applications of the MLP-Level Sets to detection of contours on synthetics and real images are presented.

Original languageEnglish
Title of host publicationMedical Imaging 2007
Subtitle of host publicationUltrasonic Imaging and Signal Processing
PublisherSPIE
ISBN (Print)081946631X, 9780819466310
DOIs
Publication statusPublished - 2007
EventMedical Imaging 2007 - San Diego, United States
Duration: 18 Feb 200722 Feb 2007

Publication series

NameProceedings of SPIE
Volume6513
ISSN (Print)1605-7422

Conference

ConferenceMedical Imaging 2007
Country/TerritoryUnited States
CitySan Diego
Period18/02/0722/02/07

Keywords

  • Level sets
  • Multilayer perceptron
  • Segmentation
  • Ultrasound images

ASJC Scopus subject areas

  • Electronic, Optical and Magnetic Materials
  • Atomic and Molecular Physics, and Optics
  • Biomaterials
  • Radiology Nuclear Medicine and imaging

Fingerprint

Dive into the research topics of 'Multilayer perceptron based level sets for robust ultrasound image segmentation'. Together they form a unique fingerprint.

Cite this