Entropy Based Large Scale Random Simulation of Brain Tumor Detection Under Monto Carlo Fuzzy Techniques

Abstract

Brain tumors are still considered one of the most challenging neurological conditions to deal with in terms of diagnosis, prognostic evaluation, and treatment because of the uncertainties and heterogeneity of this problem. Based on recent statistics, about 321,476 people have developed brain and central nervous system (CNS) tumors around the world, whereas 248,305 people died from this illness. Moreover, about 94% of patients diagnosed with glioblastoma, which is a type of malignant brain tumor, do not survive for more than five years since its five-year survival rate is less than 7%. As can be seen, the global burden of brain and CNS tumors is expected to rise sharply until 2040. It means that there is a need to develop a sophisticated approach to support decision making. This study aims at developing an integrated model based on fuzzy theory and Monte Carlo simulation in order to conduct proper analysis and optimal evaluation of the zones affected by brain tumors through processing multidimensional data obtained from different areas of the body. Using the fuzzy set theory allowed the researchers to address the uncertainties associated with medical information regarding brain tumors. Large-scale random simulations based on Monte Carlo techniques were used for creating numerous realizations. Entropy-based weighting systems were used in an effort to establish which criteria hold more diagnostic significance in comparison to others. Three types of entropy were considered in an attempt to assess stability and consistency of the prioritization process. It was revealed that the same priorities were observed across all entropy criteria with the resulting ordering being as follows (D_4 > D_3 > D_1 > D_2). From all considered options, it was clear that (D4) represented the best choice, demonstrating more significance as well as consistent results among different entropy methods of evaluation. Consistency of the results confirms the applicability of the suggested hybrid approach. Its effectiveness in establishing priority among different brain tumor zones proves that this methodology is effective in decision making related to neuro-oncology. Such hybrid framework can be successfully utilized in creating decision support systems within the brain demographic dataset for optimised selection.