TRANSFER LEARNING BASED DL MODEL OF A CROSSRESIDUAL NETWORK (XRESNET-50) WAS PRESENTED FOR RENAL KIDNEY STONE IMAGES

Abstract

The kidneys play a crucial role in human health by filtering the blood. Maintaining normal
amounts of sodium, potassium, and blood pH depends on the kidneys functioning normally. Humans are
increasingly susceptible to renal failure as a result of modern living, diet, and illnesses like diabetes. Timely
therapy of renal stones requires accurate early prognosis. The success rate of image processing-based
diagnostic methods is higher than that of other methods of identification. A DL model of a cross-residual
network (XResNet-50) was presented for renal stone categorization by Yildirim et al.2021. For precise
diagnostics, the suggested XResNet-50 uses four layers of computing. ResNet layers at every step of the
process boost the model's ability to identify features. According to the experiments, the suggested crosslayered
model has a better precision (96.23%) than the other proposed models.