让肠镜图像生成更懂病灶的描述，提升报告准确性和临床可用性。
Lesion-Aware Visual-Language Fusion for Automated Image Captioning of Ulcerative Colitis Endoscopic Examinations
- 融合热力图与临床指标，让模型关注关键病灶区域
- 在MES评分0-3范围内实现更高分类准确率
- 生成符合临床习惯的结构化描述，适合医生使用
我们提出一种针对溃疡性结肠炎（UC）的病灶感知图像描述生成框架。模型结合ResNet特征、Grad-CAM热力图与CBAM增强注意力机制，并通过T5解码器生成文本。将临床指标（MES评分0-3，血管纹理、出血、充血、易脆性、溃疡）作为自然语言提示注入，引导描述生成。系统输出结构化、可解释的描述，同时支持MES分级与病灶标签。相比基线方法，本方案提升了描述质量与MES分类准确率，助力可靠内镜报告生成。
原文摘要 · Abstract (English)
We present a lesion-aware image captioning framework for ulcerative colitis (UC). The model integrates ResNet embeddings, Grad-CAM heatmaps, and CBAM-enhanced attention with a T5 decoder. Clinical metadata (MES score 0-3, vascular pattern, bleeding, erythema, friability, ulceration) is injected as natural-language prompts to guide caption generation. The system produces structured, interpretable descriptions aligned with clinical practice and provides MES classification and lesion tags. Compared with baselines, our approach improves caption quality and MES classification accuracy, supporting reliable endoscopic reporting.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。