Extraction and segmentation of sputum cells for lung cancer early diagnosis

Fatma Taher, Naoufel Werghi, Hussain Al-Ahmad, Christian Donner

Research output: Contribution to journalArticlepeer-review

23 Scopus citations

Abstract

Lung cancer has been the largest cause of cancer deaths worldwide with an overall 5-year survival rate of only 15%. Its symptoms can be found exclusively in advanced stages where the chances for patients to survive are very low, thus making the mortality rate the highest among all other types of cancer. The present work deals with the attempt to design computer-aided detection or diagnosis (CAD) systems for early detection of lung cancer based on the analysis of sputum color images. The aim is to reduce the false negative rate and to increase the true positive rate as much as possible. The early detection of lung cancer from sputum images is a challenging problem, due to both the structure of the cancer cells and the stained method which are employed in the formulation of the sputum cells. We present here a framework for the extraction and segmentation of sputum cells in sputum images using, respectively, a threshold classifier, a Bayesian classification and mean shift segmentation. Our methods are validated and compared with other competitive techniques via a series of experimentation conducted with a data set of 100 images. The extraction and segmentation results will be used as a base for a CAD system for early detection of lung cancer which will improve the chances of survival for the patient.

Original languageBritish English
Pages (from-to)512-531
Number of pages20
JournalAlgorithms
Volume6
Issue number3
DOIs
StatePublished - 2013

Keywords

  • Bayesian classification
  • Image segmentation
  • Lung cancer detection
  • Mean shift algorithm
  • Sputum cells
  • Threshold algorithm

Fingerprint

Dive into the research topics of 'Extraction and segmentation of sputum cells for lung cancer early diagnosis'. Together they form a unique fingerprint.

Cite this