发表机构
Kyushu University（九州大学）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
针对前列腺癌组织病理学中实例级Gleason注释少致补丁级学习难的问题，提出多实例学习框架，依据Gleason评分临床定义，通过汇总实例预测等方式制定实例级学习，在SICAP-MIL数据集上表现优于现有方法。
AI 中文摘要
在前列腺癌组织病理学中，Gleason评分由全切片图像中最常见（主要）和第二常见（次要）的Gleason模式决定。虽然这些切片级标签在临床实践中常规可用，但很少提供实例级Gleason注释，这使得补丁级学习具有挑战性。我们提出了一个多实例学习（MIL）框架，该框架从切片级主要和次要标签估计实例级Gleason模式。所提出的方法根据Gleason评分的临床定义，通过将实例预测汇总为类别计数并明确建模主要模式、次要模式及其优势来制定实例级学习。实验结果表明，所提出的公式能够实现有效的实例级学习，并且在SICAP-MIL数据集上优于现有的MIL方法。
英文摘要
In prostate cancer histopathology, the Gleason Score is determined by the most frequent (Primary) and second most frequent (Secondary) Gleason patterns within a whole-slide image. Although these slide-level labels are routinely available in clinical practice, instance-level Gleason annotations are rarely provided, making patch-level learning challenging. We propose a Multiple Instance Learning (MIL) framework that estimates instance-level Gleason patterns from slide-level Primary and Secondary labels. The proposed method formulates instance-level learning according to the clinical definition of the Gleason Score by aggregating instance predictions into class counts and explicitly modeling the Primary pattern, Secondary pattern, and their dominance. Experimental results demonstrate that the proposed formulation enables effective instance-level learning and outperforms existing MIL approaches on the SICAP-MIL dataset.
CommentsAccepted to MICCAI workshop 2026 (AMAI)
