International Research Journal of Education and Technology

Due to the exponential advancement of information technology, computer, storage systems and networking technology, medical devices and clinical diagnosis has acquired tremendous popularity in the last two decades. Mostly in medical field including biomedical science, the effect of such advances is becoming apparent, allowing the detection and diagnosis in a much more vivid way. The significant hurdle in the diagnostic imaging study is to select an image without any substantial details being lost. It is extremely likely that throughout the course of retrieval or again subsequent processing phases, the data captured can be distorted by noises or artifacts. Noise is defined as the initial pixel value being modified at random. Noise lowers the clarity of the image which is particularly important whenever the structures are scanned are smaller and even have comparatively poor intensity. De-noising of image data is thus important, and in medical diagnostics it has always been a necessary pre-processing level. An analysis of several significant research in the field of image de-noising is discussed in this article. Since images were quite essential in any area, image de-noising is indeed a valuable pre-process prior towards more image analysis, such as segmentation, extraction of features, texture analysis, etc. This research intended to perform the comprehensive study of various de-noising strategies for medical imaging that involves MRI, CT and Retinal fundus images. A comparison study with many existing methods approaches focused on resemblance tests, reveals that the proposed approach is superior in image consistency to them.

Keyword: Filtering, MRI, CT Scan, Retinal, SSIM.