Hevristično generiranje medicinskih simulacijskih scenarijev : doktorska disertacija

slovenian

doktorska dela

2009

Univerza v Mariboru, Fakulteta za elektrotehniko računalništvo in informatiko

Modern education in health care today can not exist without the use medical simulators, which are state of the art for efficient learning. Simulators use scenarios, which are designed by physician's own experiences and presents a subjective approach by scenario design. In doctoral thesis machine learning methods for building scenarios from real data are presented. In particularly we introduce Bayesian probability networks as the primary tool for scenario design. We present a methodology CASGen (Computer Aided Scenario Generation) which uses Bayesian network (BN) and apply it of real study of emergency medicine, where 737 cases in prehospital resuscitation were collected. The results of CASGen are validated with the three phase model for ventricular fibrillation. We confirm the hypothesis of the research that BN are appropriate method for supporting scenario design. The application and use of CASGen was independently confirmed by experts from medical domain.

07.08.2023