Klasifikasi Penyakit Ginjal Menggunakan Convolutional Neural Network

Abstract

The kidney is an organ located at the bottom of the back rib cage. Kidney is a very important organ in the body's metabolic system. Kidney disease is associated with a risk of cardiovascular disease, which has a high treatment cost and mortality rate. Kidney failure can also occur if kidney abnormalities are not discovered and treated promptly, hence early diagnosis is an important step. Therefore, a method is needed to deal with this problem more quickly and accurately, kidney disease can be classified through CT scan images. One solution is Convolutional Neural Network (CNN), a method that uses image input that can determine the objects contained in an image so that the machine can recognize and distinguish between one image and another. Various CNN architectures have been developed. Factors that distinguish these architectures such as the number of network layers and the way the convolution layer is placed. One of them is the lightweight CNN model with its fast performance and small size but not sacrificing the much-needed accuracy performance. Lightweight CNN was used to classify the condition of the kidney through the given CT scan images. The accuracy results obtained were 98% for training data, 99% for validation data and 99% for testing data, implementated as a web-based application.