Fuzzy-logic-inspired Multi-contrast-agent Strategy for Optimal Tumor Classification

Abstract

This paper proposes a new fuzzy-logic-inspired multi-contrast-agent strategy (MCAS) for optimal tumor classification. The proposed strategy accounts for the competitive and symbiotic relationships among multiple contrast agents through a sequential logic circuit analysis. Furthermore, the strategy enables an intuitive yet systematic way to analyze the tumor classification vagueness and ambiguous uncertainties and optimize the utilization of multiple agents through a fuzzy comprehensive evaluation. A numerical example is used to demonstrate how the classification performance in terms of decision-making fuzziness is significantly improved with an optimal “cocktail recipe” methodology using the proposed MCAS.