Fast Fuzzy C Means Clustering and Adaptive Mean Shift Threshold Segmentation for Lung Cancer on CT Images

Abstract

The accurate identification of lung cancer on the CT scan is essential for the early diagnosis of it, which enables the patient with better treatment. The clustering technique has been used for presenting under computer visual impairment, but there are many challenges which continuous in detecting lung cancer because of the inadequate database. In this research, Computer Tomography the CT scan is used to scan the human organs, and to visually represent the issues that occurs in the human body. The CT scans are mostly used for detecting lung cancer and tumor, and the tracking of it is so detailed. In this research a new methodology is proposed for segmentation and classification of lung cancer using the CT scans. The CT scan images are less radiant when compared to those of MRI (magnetic resonance imaging). The result of the CT images is sometimes poor and is highly disorder due to the noise created in it. But it can be removed using the preprocessing method used in the CT scan, which also improves the quality in terms of contrast and its brightness. The 2D Adaptive Gabor Diffusion Filter (2D-AGDF) algorithm is proposed to remove various types of noises. The Edge Preserved Contrast Limited Adaptive Histogram Equalization (EP-CLAHE) algorithm is used for improving the contrast and brightness of the image from the CT scan.

The Fast-Fuzzy C Means is a hybrid clustering approach applied for the clustering of the CT image to group the CT lung cancer region. The Adaptive Mean Shift Threshold (AMST) is the methodology applied for the threshold lung cancer region in order to segment the lung cancer portion. The Experimental result demonstrates the proposed methodology is providing with improved accuracy and efficiency.