Deep deterministic policy gradient-based automatic negotiation framework for shared decision-making

Shared Decision-Making (SDM), a patient-centered approach to medical care, improves treatment outcomes and patient satisfaction. However, traditional SDM struggles in handling complex medical scenarios, dynamic patient preferences, and multi-issue negotiations, particularly under incomplete information. The key challenge lies in capturing the fuzzy preferences of doctors and patients while ensuring efficient and fair multi-issue negotiations. This study introduces AutoSDM-DDPG, an automated SDM framework based on the Deep Deterministic Policy Gradient (DDPG) algorithm. Using an Actor-Critic network, the framework dynamically optimizes negotiation strategies to address multidimensional demands and resolve preference inconsistencies in treatment planning. Fuzzy membership functions are applied to model the uncertainty in patient preferences, enhancing representation and improving multi-issue negotiation outcomes. Experimental results show that AutoSDM-DDPG outperforms other models in key indicators, including social welfare, satisfaction disparity, and decision quality. It achieves faster and more equitable negotiations while balancing the needs of both doctors and patients. In scenarios involving multi-issue negotiations and complex preferences, AutoSDM-DDPG demonstrates exceptional adaptability, achieving efficient and fair decision-making.