Prediction of Kidney Failure Using Artificial Neural

Abstract

This research intents to assess the application of artificial neural network in predicting the kidney failure disease. Kidney failure disease is being observed as a serious challenge to the field of medical with its impact on a mass population of the world. This work explored and analyzed the data generated from 60 kidney patients in many hospitals and hemodialysis centers using data mining technique . This is done by using Artificial Neural Network technique to select the weight, and connectivity structure to determine system for input variables learning. This work provides Physicians with an instrument assess the dialysis service performance The study for prediction of kidney failure has been carried out using Feed forward back propagation and Cascade forward back propagation algorithms. The results of this study demonstrate that an ANN model with variables consisting of (urea, creatinine, potassium, sodium, calcium, phosphorus and uric acid) is classified a total of 60 patients correctly to normal and abnormal. The best network model produced prediction accuracy of 98.3 percent is given by Feed Forward Back Propagation network (FFBP) .