Lung Cancer Diagnosis from Ct Images Based on Pre-Processing and Ensemble Learning

Abstract

Lung cancer is one of the dreaded diseases. Early detection is highly recommended to save the lives of the people. If the early detection is focused, we can reduce the mortality rate and increase the life expectancy accuracy 13% of the all new cancer diagnosis and 24% of all cancer deaths. There are various methods to detect the lung cancer from x-rays and CT images but the CT images are preferred. The medical images are always preferred to get better results on the same disease. The proposed method here will discuss about how the Watershed segmentation can be applied to the pre-processing of CT scans. The results are used for the deep learning methods so as to get the more accuracy from more image scans. The results are then used for the verification by the medical examiner for the validity of the results. The major pre-processing is done by using Median Filter, Gabor Filter and Watershed segmentation. In this research work we will discuss how the image manipulation can be done to achieve better results from the CT images through various image processing methods. The construction of the proposed method will include smoothing of the images with median filters, enhancement of the image and finally segmentation of the images.