Denoising Technique and Analysis of Statistical Parameters for Endoscopic Images of Gastric Cancer

Gastric cancer (GC), which is commonly referred to as stomach cancer, is a kind of cancer that develops in the stomach cells. The gastrointestinal tract is a component of the alimentary canal, which is composed of a series of spherical muscular organs linked by a long, curved tube that passes from the inside of the mouth to the anus. Endoscopic gastroduodenoscopy (EGD) using an upper endoscope is a procedure that helps identify most stomach cancers because the endoscopic pictures that using filtering algorithm to increase endoscopic image tissue that are obtained vulnerable noises. Denoising is extremely important of the endoscopic images in diagnosis and subsequent medical care. On the basis of the endoscopic imaging data that has been obtained, filtering and denoising for image management were performed in order to obtain quality data and, as a result, to protect and safeguard the dependability of important clinical information. During the course of this study, numerous algorithms like Gaussian, median, wavelet, Weiner, filtering were performed on the selected endoscopic images and the resulting data has been statistically analyzed and characterized in detail. The various filtering algorithms were used on the endoscopic images and the results have been analyzed using a descriptive parameter statistical analysis approach. The skewness and kurtosis values for all the filtered images were calculated and compared. The results obtained from the parametric method were analyzed using various tools like ANOVA, t-test, z-test, linear regression, and covariance. The data obtained were used to calculate the ranking and percentile for analyzing the performance of the filters. Based on the results obtained from the analysis the good and best filter that can be used for denoising endoscopic images of gastric cancer was identified.