Personalized surgical treatment recommendation with joint consideration of multiple decision-making dimensions

Abstract Surgical treatment planning is a highly complicated and personalized procedure, where a surgeon needs to balance multiple decision-making dimensions including effectiveness, risk, and cost wisely for the best benefit of the patient, based on his/her conditions and preferences. Developing an algorithm-driven support system for surgical treatment planning is a great appeal. This study fills in this gap with MUBA (multidimensional Bayesian recommendation), an interpretable data-driven intelligent system that supports personalized surgical treatment recommendations on both the patient’s and the surgeon’s side with joint consideration of multiple decision-making dimensions. Applied to surgical treatment recommendation for Pelvic Organ Prolapse, a common female disease with significant negative impacts on the life quality of patients, MUBA achieved excellent performance that was comparable to top urogynecologists, with a transparent decision-making process that made communications between surgeons and patients much easier. Such a success indicates that MUBA has good potential in solving similar problems in other diseases.