FG-CXR：一个放射科医生对齐的注视数据集，用于增强胸部X光报告生成的可解释性
FG-CXR: A Radiologist-Aligned Gaze Dataset for Enhancing Interpretability in Chest X-Ray Report Generation
- University of Arkansas（阿肯色大学）
- West Virginia University（西弗吉尼亚大学）
- University of Liverpool（利物浦大学）
- MD Anderson Cancer Center（MD安德森癌症中心）
- University of Houston（休斯顿大学）
机构由 AI 辅助整理，请以论文原文为准。
AI总结：
本文提出FG-CXR细粒度注视数据集和可解释注意力生成网络Gen-XAI，通过将生成报告与放射科医生的注视注意力和诊断文本对齐，增强胸部X光报告生成的可解释性。
AI中文摘要：
在胸部X光（CXR）分析中开发可解释的报告生成系统，在计算机辅助诊断（CAD）系统中正变得越来越关键，它使放射科医生能够理解这些系统所做的决策。尽管关注报告生成的多样化数据集和方法不断增长，但这些模型生成的报告与真实放射科医生的解读之间的对齐程度仍存在显著差距。在本研究中，我们通过首先引入细粒度CXR（FG-CXR）数据集来应对这一挑战，该数据集提供了放射科医生生成的描述与每个解剖结构对应的注视注意力热图之间的细粒度配对信息。与现有数据集包含原始注视序列和报告、且注视位置与报告内容之间存在显著不对齐不同，我们的FG-CXR数据集在注视注意力与诊断文本之间提供了更细粒度的对齐。此外，我们的分析表明，简单地应用黑箱图像描述方法来生成报告，无法充分解释CXR中哪些信息被利用以及需要关注多长时间才能准确生成报告。因此，我们提出了一种新颖的可解释放射科医生注意力生成网络（Gen-XAI），该网络模拟放射科医生的诊断过程，明确约束其输出以同时紧密对齐放射科医生的注视注意力和文本记录。最后，我们进行了大量实验以说明我们方法的有效性。我们的数据集和检查点可在https://github.com/UARK-AICV/FG-CXR获取。
英文摘要：
Developing an interpretable system for generating reports in chest X-ray (CXR) analysis is becoming increasingly crucial in Computer-aided Diagnosis (CAD) systems, enabling radiologists to comprehend the decisions made by these systems. Despite the growth of diverse datasets and methods focusing on report generation, there remains a notable gap in how closely these models' generated reports align with the interpretations of real radiologists. In this study, we tackle this challenge by initially introducing Fine-Grained CXR (FG-CXR) dataset, which provides fine-grained paired information between the captions generated by radiologists and the corresponding gaze attention heatmaps for each anatomy. Unlike existing datasets that include a raw sequence of gaze alongside a report, with significant misalignment between gaze location and report content, our FG-CXR dataset offers a more grained alignment between gaze attention and diagnosis transcript. Furthermore, our analysis reveals that simply applying black-box image captioning methods to generate reports cannot adequately explain which information in CXR is utilized and how long needs to attend to accurately generate reports. Consequently, we propose a novel explainable radiologist's attention generator network (Gen-XAI) that mimics the diagnosis process of radiologists, explicitly constraining its output to closely align with both radiologist's gaze attention and transcript. Finally, we perform extensive experiments to illustrate the effectiveness of our method. Our datasets and checkpoint is available at https://github.com/UARK-AICV/FG-CXR.
