Towards an Intelligent Data Analysis System for Decision Making in Medical Diagnostics

Artificial neural networks (ANN) are currently massively used in different fields, especially for very complex problems. In this work we propose an approach to use these systems, and in particular the paradigm of the self-organizing map (SOM) in the medical field. The idea is to use this paradigm to develop an intelligent system able of learning to analyze, classify, and visualize multi-parameter objects in a reduced two-dimensional space in the form of object maps. This approach allows for the visual analysis and interpretation of data to reveal the most informative indicators for decision making. The application in the medical field aims to help make a very good diagnosis to make the most relevant decisions in order to provide appropriate treatment depending on the patient’s state.