Detection of Gallbladder Disorder through Iris Using Deep Learning Models for Classification

Abstract

An analysis of the distinct patterns present in the iris enables the identification of possible markers of gallbladder problems via the application of deep learning. Deep learning techniques are a dependable and quickly developing field for diagnosing disorders, as demonstrated by biomedical research. This study uses human iris to assess the diagnostic accuracy of iridology for gallbladder issues. The training dataset consists of 2481 images belonging to two classes with normal and abnormal gallbladder was used in the study. The iris images, each measuring 640 by 480 pixels, were taken. Using the iridology chart as a reference, the focal area in the iris image was chosen so that it matched the location of the specific organ. Then, using statistical analysis features and the 2D Discrete Wavelet Transform (DWT), characteristics from this region features were extracted. The results demonstrate that the classifier generated the highest classification accuracy by using ResNet-50 It has been demonstrated that the suggested model for the automatic and non-invasive detection of gallbladder disorders is both diagnostically significant and effective.