A Survey on Detection and Classification of Chronic Kidney Disease with a Machine Learning Algorithm

Abstract

The primary goal of this study is to detect and diagnose chronic kidney diseases (CKDs), specifically kidney stones, cystic kidneydisease, and suspected renal carcinoma. CKDs pave the way for a variety of diseases that aren't related to the urinal system.Coronary heart disease, stroke, cardiomyopathy, pulmonary hypertension, and heart valve disease will allbecome more likely asa result of it. Early diagnosis of chronickidney disease will savelivesand avoid theonset of more severe illnesses. Forabdominal research, ultrasound imaging is a commonly used diagnostic tool. Chronic kidney diseases were identified using aframework that included a Histogram of directed gradient function and the KNN Algorithm in this proposed method. The multi-layered Convolution Neural Network (CNN) architecture hasbeen trained for kidney disease classification, and the Batchprediction approach has been tested for chronic kidney disease forecasting. The detection accuracy for kidney disease is reportedtobe96.67percent.Theclassificationof CKDultrasoundusingCNN hasanaccuracyof85.2percent.