A computer science discipline based on the adaptation of the system using training information. Using a learning procedure, a mathematical mechanism changes its parameters to recognise the features that represent the patterns of a problem (and discrimination between classes) by using examples. Pattern Recognition has its application in many types of problem, where a classification or regression solution is required and the model of the problem is difficult to obtain. The inductive approximation allows the system to learn the important features of the cases to make a classification with different classes defined by the values of the variable or character.

The practical result of the pattern recognition experiments should be a clinical decision support system to improve the quality in the clinical decision. Defined as an active knowledge system, generating specific advice for each new case, it will integrate the three principal features of these systems (see figure below): Medical knowledge that is used to solve the disease problems for the cases; Patient data: specific biomedical information of each patient; and Specific advice for each case: based on the medical knowledge and the patient data, the system generates a specific result.

This new methodology will have a direct application for better diagnosis, prognosis and treatment selection of brein tumours, with a positive impact in the health of patients.

.

Pattern Recognition Methodology

Experts within the consortium  ITACA and Katholieke Universiteit Leuven