Yapay zeka teknikleri ile tıbbi karar destek uygulamaları

Medical science is very receptive to the support and integration of technological developments, and artificial intelligence has a major role in technological advancements of this era. The purpose of this thesis is to build frameworks and establish mathematical models that will assist decision makers (physicians) in critical decision moments by using artificial intelligence techniques of engineering in medical decisions. Within the scope, three diseases are included: Diabetes, Tuberculosis and Chronic Obstructive Pulmonary Disease (COPD). All three of them are reported among top ten global causes of death by World Health Organization (WHO), recently in the global health statistics. In order to contribute to the management of these diseases by establishing medical support systems, four different artificial intelligence techniques are used. Fuzzy C-means algorithm is used for diagnosis of diabetes, intuitionistic fuzzy cognitive maps, which are found effective in social sciences as marketing, are used for detection of multi-drug resistant tuberculosis, the novel statistical-based fuzzy cognitive map is proposed for the management of COPD and, artificial neural networks are used to predict the length of hospital stay of COPD patients. Each section of this thesis is an independent application of different artificial intelligence techniques, constructing medical decision support frameworks. According to the numerical applications, the most successful results are obtained using the approaches that combine the powers of artificial intelligence and human knowledge.