Diagnosis of Chest Diseases Using Artificial Neural Networks

Abstract

The important health problems in the world are mainly caused due to chest diseases. A comparative chest disease diagnosis has been realized in this study by using the following neural networks such as multi-layer, probabilistic, learning vector quantization, constructive fuzzy, focused time delay, and generalized regression neural networks [1]. The back-propagation algorithm (It is a method used to calculate the error contribution of each neuron after a batch of data has been processed. It is the workhorse of learning in neural networks.) is the most popular algorithm in feed-forward neural network [1] with a multi-layer system. By adjusting the weights of artificial neural network while moving along the descending gradient direction is applicable to calculate the output error as well as the gradient of the error. The theme is to propose the implementation of back-propagation algorithm to compute and compare the percentage of the output accuracy, which is used for medical diagnosis on various chest diseases (i.e., asthma [2], tuberculosis [1, 3], lung cancer [4]; it is an iterative procedure that generates pneumonia).