A Comparative Analysis of Various Filters to Denoise Medical X-Ray Images

Abstract

in recent years, processing of medical images and analysing them is an evolving field for researchers across the globe. Various medical imaging methods like, Computed Tomography scans (CT scans), and Magnetic Resonance Imaging (MRI), X-rays etc. allow medical practitioners to diagnose various abnormalities in the human body. X ray images play vital role in medical fields for detection of various diseases. But, due to the presence of various noises, important medical information can be corrupted and altered, which surely can lead to incorrect diagnosis. Though advanced medical imaging technologies are being used, still the effect of noise could not be eradicated fully. Moreover, with various medical image processing technologies, medical X-Ray images are being used for automatic detection of various diseases, in such cases also, noises present in the image can affect the accuracy of the learning models. In this research artificial noise modeling was applied to medical X-Ray images, and a comparative analysis was done to find which filtering scheme works better for which type of noises, on the basis of various parameters like MSE, PSNR, SNR, and SSIM. It was found that, for both kind of noises, HMF Filtering is quite effective. Anisotropic Diffusion Filtering also performed satisfactory. Though, no filter could be declared as most effective on both type of noises. But, their performance varied according to type of noise present and evaluation parameters.