Open AccessFuzzy C-Means and Antlion Optimization Based Segmentation of Juxtapleural Lung Nodules
Author(s) -
Parvathi Kishore P,
R. Rajeswari
Publication year - 2020
Publication title -
international journal of recent technology and engineering
Language(s) - English
Resource type - Journals
ISSN - 2277-3878
DOI - 10.35940/ijrte.e6270.018520
Subject(s) - fuzzy logic , segmentation , lung , cluster analysis , computed tomography , nodule (geology) , artificial intelligence , computer science , silhouette , lung cancer , entropy (arrow of time) , pattern recognition (psychology) , mathematics , radiology , biology , medicine , pathology , physics , paleontology , quantum mechanics
A Computer aided detection of lung nodules plays a vital role in diagnosis of lung cancer. The aim of this paper is to utilize the characteristics of hybrid Fuzzy C-Means-Ant lion Optimization (FCM-ALO) and morphological operations to extract the juxtapleural lung nodules. The hybrid FCM-ALO based clustering helps in isolating the nodules and boundaries of lung lobes. Morphological operations are then applied to isolate the juxtapleural nodules from the lung boundaries. The proposed method is evaluated using 28 computed tomography (CT) case studies from Lung Imaging Database Consortium-Image Database Resource Initiative (LIDC-IDRI) with 100 juxtapleural nodules. The FCM-ALO based clustering approach gives 0.9464, 0.1575 and 0.2009 as average silhouette index (S), Davies-Bouldin index (DB) and entropy respectively. The sensitivity, specificity and accuracy of the proposed juxtapleural lung nodule segmentation are 99.5\%, 95.03%, and 97.63% respectively.