Risk assessment of coronary arteries heart disease based on neuro-fuzzy classifiers

Coronary artery heart disease is one of the main reasons of death in under development countries such as Iran. Based on vagueness in data and uncertainty in decision making finding an optimal way for diagnosis would be helpful. The main goal of this article is to change the linguistic terms which doctors use for representing coronary heart disease possibility into classified stages. Indeed the medical diagnosis can be used as an issue of pattern recognition by an input vector of features introduced to a system so the corresponding risk of a particular disease can be estimated with this system. In this paper we used an adaptive network-based inference system in order to assess the risk of coronary artery heart diseases .With this point of view that fault in medical diagnosis are more errors of omission than of commission, we are going to reduce this diagnosis defect with Neuro-fuzzy networks such as Multi-Layer Perceptron (MLP) and an inference Neuro-fuzzy network such as ANFIS. In the last section we will compare these methods to show that ANFIS can obtain the best accommodation with doctor's opinions.