文献类型：会议论文

英文题名：Radiology Report Generation via Multi-Objective Preference Optimization

作者：Xiao, Ting[1];Shi, Lei[1];Liu, Peng[2];Wang, Zhe[1];Bai, Chenjia[3]

机构：[1]East China Univ Sci & Technol, Shanghai, Peoples R China;[2]Harbin Inst Technol, Harbin, Peoples R China;[3]China Telecom, Inst Artificial Intelligence TeleAI, Beijing, Peoples R China

会议论文集：39th AAAI Conference on Artificial Intelligence

会议日期：FEB 25-MAR 04, 2025

会议地点：Philadelphia, PA

语种：英文

摘要：Automatic Radiology Report Generation (RRG) is an important topic for alleviating the substantial workload of radiologists. Existing RRG approaches rely on supervised regression based on different architectures or additional knowledge injection, while the generated report may not align optimally with radiologists' preferences. Especially, since the preferences of radiologists are inherently heterogeneous and multi-dimensional, e.g., some may prioritize report fluency, while others emphasize clinical accuracy. To address this problem, we propose a new RRG method via Multi-objective Preference Optimization (MPO) to align the pre-trained RRG model with multiple human preferences, which can be formulated by multi-dimensional reward functions and optimized by multi-objective reinforcement learning (RL). Specifically, we use a preference vector to represent the weight of preferences and use it as a condition for the RRG model. Then, a linearly weighed reward is obtained via a dot product between the preference vector and multi-dimensional reward. Next, the RRG model is optimized to align with the preference vector by optimizing such a reward via RL. In the training stage, we randomly sample diverse preference vectors from the preference space and align the model by optimizing the weighted multi-objective rewards, which leads to an optimal policy on the entire preference space. When inference, our model can generate reports aligned with specific preferences without further fine-tuning. Extensive experiments on two public datasets show the proposed method can generate reports that cater to different preferences in a single model and achieve state-of-the-art performance.