Title: Nodule detection from posterior and anterior chest radiographs with different methods
Abstract: Lung cancer is the deadliest type of cancer in the world, for both men and women. Hence early detection is the most promising way to improve the patient's chance for survival from lung cancer. The most common technique used to examine the lung cancer is Posterior and Anterior chest radiography and computerized tomography scans. PA chest radiography is the cost effective tool in diagnosing lung tumors. But interpreting chest radio graph is difficult because of superimposed anatomical structure present in the image. Even experienced radiologist facing difficult in finding abnormalities in the X-ray image. Different Computer Aided Diagnosis system proposed in the literature to help the radiologists in finding abnormalities from X-Ray images, but none showed perfect result because of complex anatomical structure present in the image and subtlety of the nodule. In this paper three methods are proposed to find tumor from PA chest radio graphic images. All the three methods are developed and tested on standard database available on Japanese society of radiological Technology and images available on web. In the first method tumor part delineated from the image with the help of different morphological operations and mask, then various features are obtained from the segmented tumor. In the second method suspicious nodule regions are found by applying multi scale blob detection method and to reduce number of false positives threshold method is employed. For further reduction of false positives and to detect potential nodule SVM classifier is used. In the third method lung part of the image is divided in to patches of fixed size. Patch contain potential nodule found by using Restricted Boltzmann machine and SVM classifier. Here Restricted Boltzmann is used for finding the features from patch (region of interest), ROI is classified as nodule (or) non-nodule by using linear SVM classifier.
Publication Year: 2016
Publication Date: 2016-12-01
Language: en
Type: article
Indexed In: ['crossref']
Access and Citation
Cited By Count: 1
AI Researcher Chatbot
Get quick answers to your questions about the article from our AI researcher chatbot