Application of the Fuzzy Min-Max Neural Networks to Medical Diagnosis

Authors:
Anas Quteishat;Chee Peng Lim
Affiliations:
School of Electrical & Electronic Engineering, University Science Malaysia, Malaysia;School of Electrical & Electronic Engineering, University Science Malaysia, Malaysia
Venue:
KES '08 Proceedings of the 12th international conference on Knowledge-Based Intelligent Information and Engineering Systems, Part III
Year:
2008

Citing 1
Cited 1

A modified fuzzy min-max neural network with rule extraction and its application to fault detection and classification

Applied Soft Computing

Medical pattern recognition: applying an improved intuitionistic fuzzy cross-entropy approach

Advances in Fuzzy Systems - Special issue on Hybrid Biomedical Intelligent Systems

Quantified Score

Hi-index	0.00

Visualization

Abstract

In this paper, the Fuzzy Min-Max (FMM) neural network along with two modified FMM models are used for tackling medical diagnostic problems. The original FMM network establishes hyperboxes with fuzzy sets in its structure for classifying input patterns into different output categories. While the first modified FMM model uses the membership function and the Euclidian distance to classify the input patterns, the second modified FMM model employs only the Euclidian distance for the same process. Unlike the original FMM network, the two modified FMM models undergo a pruning process, after network training, to remove hyperboxes with low confidence factors. To assess the effectiveness of the three FMM networks in medical diagnosis, a set of real medical records from suspected Acute Coronary Syndrome (ACS) patients is collected and used for experimentation. The bootstrap method is used to analyze the results statistically. Implications of the experimental outcomes are discussed, and the potential of using the FMM networks a decision support tool for medical prognostic and diagnostic problems is demonstrated.