arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2304.09185cs.CLcs.AI
收藏

面向放射学报告生成的 Token 不平衡自适应

Token Imbalance Adaptation for Radiology Report Generation

University of Memphis(孟菲斯大学);St. Jude Children’s Research Hospital(圣裘德儿童研究医院)

Yuexin Wu, I-Chan Huang, Xiaolei Huang

AI总结:

针对放射学报告中低频医学 token 生成不足的问题,提出 TIMER,利用非似然损失和强化学习动态增强低频 token,在 IU X-RAY 与 MIMIC-CXR 上提升了鲁棒性。

摘要:

文本中天然存在不平衡的 token 分布,导致神经语言模型在高频 token 上过拟合。Token 不平衡可能削弱放射学报告生成器的鲁棒性,因为复杂的医学术语出现频率较低,却承载更多医学信息。本研究中,我们在放射学报告生成的两个标准基准数据集 IU X-RAY 和 MIMIC-CXR 上展示了当前最先进模型为何无法生成低频 token。为解决这一挑战,我们提出 Token Imbalance Adapater(TIMER),旨在提升低频 token 的生成鲁棒性。该模型通过非似然损失自动利用 token 不平衡,并动态优化生成过程以增强低频 token。我们在两个基准上将该方法与多种最先进方法进行比较。实验证明,本方法在整体鲁棒性和低频 token 方面均有效。消融分析表明,我们的强化学习方法在为放射学报告生成自适应 token 不平衡方面发挥了主要作用。

英文摘要:

Imbalanced token distributions naturally exist in text documents, leading neural language models to overfit on frequent tokens. The token imbalance may dampen the robustness of radiology report generators, as complex medical terms appear less frequently but reflect more medical information. In this study, we demonstrate how current state-of-the-art models fail to generate infrequent tokens on two standard benchmark datasets (IU X-RAY and MIMIC-CXR) of radiology report generation. % However, no prior study has proposed methods to adapt infrequent tokens for text generators feeding with medical images. To solve the challenge, we propose the \textbf{T}oken \textbf{Im}balance Adapt\textbf{er} (\textit{TIMER}), aiming to improve generation robustness on infrequent tokens. The model automatically leverages token imbalance by an unlikelihood loss and dynamically optimizes generation processes to augment infrequent tokens. We compare our approach with multiple state-of-the-art methods on the two benchmarks. Experiments demonstrate the effectiveness of our approach in enhancing model robustness overall and infrequent tokens. Our ablation analysis shows that our reinforcement learning method has a major effect in adapting token imbalance for radiology report generation.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑