Main Article Content

Abstract

Segmenting lung fields from CXRs (Chest Radiographs) is an important task for the analysis, diagnosis and treatment of tumor diseases. Although many segmentation method have been presented. This paper proposed for the CXRs and also for lung field in Saliency Map by using Convolution Neural Network and Thresholding for lungs. In our convolutional neural network to achieve feature extraction and classification using Machine Learning Algorithm also reduce time consumption, deliver better segmentation. In this study, we implemented a Fully Convolutional Network to simultaneously segment multiple structures, namely lung field in standard Posterior-Anterior chest radiographs. We have developed a multi scale CNN approach for segmenting lung tumors which enables accurate measurement of tumor volumes in the lung. And we are executing our method via ordinary laptop. In this method reducing time consumption, give clear shape of tumor level in lung and efficiency. Achieve very higher accuracy on most of the structure when compared with state of other segmentation method.

Article Details

How to Cite
M.Manimegalai, R.Suruthi, M.Usha, G.Vijayalakshmi, & K.Vijayalakshmi. (2021). Segmentation of lung field from saliency maps by convolution neural network in chest radiographs using machine learning. International Journal of Intellectual Advancements and Research in Engineering Computations, 7(1), 1069–1076. Retrieved from https://ijiarec.com/ijiarec/article/view/190