Abstracto

Detection and Classification of Kidney Disease Using Convolutional Neural Networks

Shaikh Abdul Hannan, Pushparaj Pal

Chronic kidney disease is a serious, long lasting condition that can be welcomed on by kidney harm or renal disease. In the present status of study, the occurrence of chronic kidney disease (CKD) rises yearly and the numbers of patients are increasing in day to day life because of bad habit and outside food. The capacity of AI calculations to group information with high exactness makes them more vital in clinical determination and a hotspot for future treatment in CKD guess. In the recent past, the effectiveness of feature selection techniques for reducing the amount of the data has been linked to the accuracy of classification algorithms. The batch prediction approach is tested for the detection and classification of chronic kidney diseases in this suggested system's multi-layered convolution neural network (CNN) architecture, which has been trained for the categorization of kidney disorders. The accuracy rate for CNN's classification of CKD ultrasonography is 87.4%. Our inventive model delivered empowering results; they work on the determination of patient circumstances with high precision, decreasing radiologist weight and giving them an instrument that can naturally evaluate kidney condition, bringing down possibility of misdiagnosis. Moreover, working on the norm of clinical consideration and early revelation can change the direction of the disease and drag out the patient's life. The aim of this paper is to detect and classify kidney disease using convolution neural network.