Computer Aided Diagnosis System for Chronic Obstructive Pulmonary Disease from CT Images Using Convolutional Neural Network

Abstract

The Chronic Obstructive Pulmonary Disease (COPD) is a kind of popular chronic disease. The main factors that cause COPD disease are inhaling dust, shortness of breath, environment pollution, fatigue and frequent respiratory infections. COPD is characterized by airflow limitation resulting from chronic inflammatory responses in the airways and noxious particles or gases. Computer Aided Diagnosis System assists the doctors for interpretation of medical images. The Computer Aided Diagnosis System has been intended to diagnosis the COPD by using CT images. Computed Tomography (CT) images are generally chosen due to less distortion, less time consumption and low cost. The proposed work of computerized based diagnosis system for a Chronic Obstructive Pulmonary Disease (COPD) is to diagnosis the disease with accuracy using convolutional neural network (CNN). CNN classifier to classify the CT images and it will be evaluated using performance metrics. The Computer Aided Diagnosis System for COPD is composed of preprocessing, feature extraction, segmentation and classification. The preprocessing is to improve the quality of image like removing noise and isolating region of interest. The feature extraction is a method of capturing visual content of images for indexing and retrieval. The segmentation subdivides the CT image into different regions. The CNN classifier is to classify the segmented CT images for improving the accuracy of clustering under noise.