Computer based medical systems

Abstract

NIH Working Definition of Bioengineering form July 1997 defines bioengineering as a discipline that integrates physical, chemical, mathematical, and computational sciences and engineering principles study biology, medicine, behavior, and health. It advances fundamental concepts; creates knowledge from the molecular the organ systems levels; and develops innovative biologics, materials, processes, implants, devices, and informatics approaches for the prevention, diagnosis, and treatment of disease, for patient rehabilitation, and for improving health. Computer based medical systems are an integral part of the above concept and in the words of a prominent researcher, Ralph Grams, they will revolutionize the ways in which we take care of patients. His words physical examination room will be transformed into a multimedia teaching studio, where physicians can pack into a brief exam a great deal of useful medical information that patients and their families can leave with, are becoming more and more realistic. The field of computer based medical systems is now active for about 20 years. In the mid1980s the IEEE Computer society established a technical committee on Computational medicine. One of the team's main tasks was advance the support medical and health working processes with information and computer technology. Many problems have been solved already but much more still remains be solved. In this special issue we present some of the novel approaches which are or will shortly be used in everyday medical practice in the manner of the popular catchphrase to improve and maintain health of all in a more efficient and effective way. In their paper Estevez, Alay6n, Moreno, Sigut and AguiLar present the design of a pattern recognition system based on a Fuzzy Finite State Machine (FFSM). They try find an optimal FFSM able recognise different patterns with Genetic Algorithms and apply it a real problem: evaluating the feasibility of a homotopic texture measurement of the chromatin distribution in cellular nuclei. Kl~ma, Kubaifk and Lhotska claim that in medical systems it is often advantageous utilize specific problem situations (cases) in addition or instead of a general model. In their paper they discuss the issues of automated tuning in order obtain a proper definition of mutual case similarity in a specific medical domain. The two case studies mortality prediction after cardiological intervention, and resource allocation at a spa document that the optimization process is influenced by various characteristics of the problem domain. Sprogar, Sprogar and Cotnari~ present an autonomous evolutionary algorithm for the construction of decision trees. It is interesting that their algorithm requires no or just minimal human interaction. It is based on a non-standard implicit fitness evaluation in the selection phase of a co-evoLving environment. Together with selfadaptation of evolution parameters and with some other improvements it can monitor and adjust its own behavior. Podgoretec, Koko[, Molan Stigtic, Heri~.ko and Rozman study an evolutionary machine learning approach data mining and know[edge discovery based on the classification rules induction. A method for automatic rules induction called AREX using evolutionary induction of decision trees and automatic programming is introduced. The proposed algorithm is applied a cardiovascular dataset consisting of different groups of attributes, which should possibly reveal the presence of some specific cardiovascular problems in young patients.