Noise Removal Techniques for Lung Cancer CT Images

Abstract

Objectives: To analyze various filtering methods to eliminate noises present in the lung CT images and to enhance the image, which help in further evaluation of CT images for accurate lung cancer detection. To compare the proposed method with existing filtering techniques and to find the best filtering technique. Methods: For input lung CT images noises like salt along with pepper noise and speckle noise are added. For noisy images different filtering methods like Median filter, Wiener filter, Gaussian filter and Guided filter are applied. The performances of different filters are computed in terms of metrics for evaluation like PSNR, SSIM, MSE, and SNR. Based on the performance metrics the best filter is selected to remove noise in the lung CT images. Findings: The results of the experiment shows that the median filter is more efficient in comparison to other filtering methods in eliminating noises that exist in lung CT images by owning fewer mean square error (MSE) value of 4.065604, a high SNR value of 36.5931, a high SSIM value of 0.983545, and high PSNR value of 42.0395. Novelty: Different filtering methods are analyzed for different noise densities from 5% to 50% and chosen best filter by considering different evaluation metrics. The proposed method is compared with existing filtering techniques. The method can be used for elimination of noise in the other imaging modalities. Keywords: Filtering; Median filter; Wiener filter; Gaussian filter; Salt and Pepper noise; Speckle noise