Social network trust relationship environment based advanced ovarian cancer treatment decision-making model: An approach based on linguistic information with experts’ multiple confidence levels

Abstract

Advanced ovarian cancer is the deadliest gynecologic cancer in World. The survival rate is not significant, and it depends on how well cancer responds to treatment. Generally, the treatment decision depends on Gynecologic oncology experts, which treatment is best for the patient. However, a Gynecologic oncology expert’s limited rationality and personality traits strongly impact a patient’s treatment decision. In addition, according to the American Cancer Society, a patient’s treatment decision may change if Gynecologic oncology experts change. To solve this kind of medical problem, we design a group decision-making model (GDM) based on linguistic preference relations (LPRs) with multiple confidence levels like confidence and doubting levels of experts to analyse experts’ treatment decision opinions digitally for improving the survival rate of patients. In a GDM, consistency analysis and obtaining expert weights are two keys. Hence, this paper design an LPRs with confidence and doubting levels (LPRs-CDLs) based GDM model under a unique consistency improvement algorithm and Pythagorean linguistic trust relations social network analysis driven expert weight determination. First, we define the additive consistency of LPRs-CDLs based on minimum confidence and maximum doubting levels. Second, a consistency reaching process (CRP) algorithm with convergence theorem is offered to improve the consistency of experts’ opinion decisions. Third, develop a trust propagation model based on Pythagorean linguistic trust relations under a social network environment for expert’s weights determination. Then, a social network trust relationship-based GDM model with LPRs-CDLs information is presented to the advanced ovarian cancer patient’s treatment final decision. Finally, we show the rationality of this study with some comparative analysis.