Segmentation & Threshoilding of Covid-19 and Lung Cancer Using Enhanced Ct Images

Abstract

Diagnosing the lung severity diseases like COVID-19 and Lung Cancer segmentation using Computed Tomography (CT) scans plays a significant role. Detecting them in the initial phase is significant to bring down the rate of mortality risk of patients. There are 2 important factors that affects the existing methods which use CT Images employ i)huge bundles for images for training ii)handling fixed network  for the completion of training process. In this research, a cascaded system is proposed to segment the lung, detect, localize, and quantify COVID-19 infections and Lung cancer from computed tomography images. The proposed system clusters and localizes infections of varied shapes and sizes. In this research, the enhanced CT scans are used and segmented using 2D Hybrid Fuzzy C Means (2D HFCM) for segmentation of lungs caused by COVID-19 as well as lung cancer. The 2D Adaptive Otsu Thresholding (2D AOT) is used for image thresholding and a comparison has been made between traditional algorithms to show the efficiency of the proposed algorithms. Further Superimposed principle is applied to show the exact location of infection.