Skip to main navigation Skip to search Skip to main content

The ROIs segmentation method of the lungs based on adaptive EM algorithm and edge gradient information

  • Ru Liu
  • , Yang Liu*
  • , Maozu Guo
  • , Rulin Ma
  • , Ping Li
  • *Corresponding author for this work
  • Harbin Institute of Technology
  • Harbin Medical University

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

Abstract

The accurate segmentation of lung nodules plays a key role in evaluating therapeutic effect in clinic. In this paper, we propose a precise pixel-level segmentation method for region of interest (ROI) in order to accurately segment lung nodules further. Firstly, we employ expectation maximization (EM) algorithm based on adaptive threshold for binarizing lung image to segment ROIs preliminarily in lung parenchyma area extracted by region growing. In addition, region boundary is detected and expanded constantly combined with edge gradient information until the approximate real boundary of ROI is acquired. The experimental data is from the LIDC public database and the database of the Second Affiliated Hospital of Harbin Medical University. Experimental results show that our approach is feasible and effective for precise extraction of ROIs, which is very helpful to segment lung nodules and reduce the false positives.

Original languageEnglish
Title of host publicationProceedings of the 2012 International Conference on Image Processing, Computer Vision, and Pattern Recognition, IPCV 2012
Pages52-57
Number of pages6
StatePublished - 2012
Event2012 International Conference on Image Processing, Computer Vision, and Pattern Recognition, IPCV 2012 - Las Vegas, NV, United States
Duration: 16 Jul 201219 Jul 2012

Publication series

NameProceedings of the 2012 International Conference on Image Processing, Computer Vision, and Pattern Recognition, IPCV 2012
Volume1

Conference

Conference2012 International Conference on Image Processing, Computer Vision, and Pattern Recognition, IPCV 2012
Country/TerritoryUnited States
CityLas Vegas, NV
Period16/07/1219/07/12

Keywords

  • EM
  • Gradient information
  • ROIs
  • Segmentation

Fingerprint

Dive into the research topics of 'The ROIs segmentation method of the lungs based on adaptive EM algorithm and edge gradient information'. Together they form a unique fingerprint.

Cite this