用于计算病理学中无监督幻灯片表示学习的形态原型
Morphological Prototyping for Unsupervised Slide Representation Learning in Computational Pathology
AI总结：
提出PANTHER，一种基于高斯混合模型和形态原型的无监督WSI表示方法，在13个数据集的亚型分类与生存任务中达到或超过有监督MIL基线。
摘要：
病理全幻灯片图像（WSIs）的表示学习主要依赖于多实例学习（MIL）的弱监督。然而，这种方法产生的幻灯片表示高度针对特定临床任务，这限制了它们的表达能力和泛化能力，尤其是在数据有限的情况下。相反，我们假设可以利用组织中的形态冗余，以无监督的方式构建与任务无关的幻灯片表示。为此，我们提出了PANTHER，这是一种基于高斯混合模型的原型方法，可将WSI图像块集合总结为一组小得多的形态原型。具体而言，假设每个图像块均由混合分布生成，其中每个混合成分代表一个形态范例。然后，利用估计的混合参数，我们构建了一个紧凑的幻灯片表示，可轻松用于广泛的下游任务。通过在13个数据集上对PANTHER在亚型分类和生存任务上进行广泛评估，我们表明：1）PANTHER优于或与有监督MIL基线相当；2）对形态原型的分析为模型可解释性带来了新的定性和定量见解。
英文摘要：
Representation learning of pathology whole-slide images (WSIs) has been has primarily relied on weak supervision with Multiple Instance Learning (MIL). However, the slide representations resulting from this approach are highly tailored to specific clinical tasks, which limits their expressivity and generalization, particularly in scenarios with limited data. Instead, we hypothesize that morphological redundancy in tissue can be leveraged to build a task-agnostic slide representation in an unsupervised fashion. To this end, we introduce PANTHER, a prototype-based approach rooted in the Gaussian mixture model that summarizes the set of WSI patches into a much smaller set of morphological prototypes. Specifically, each patch is assumed to have been generated from a mixture distribution, where each mixture component represents a morphological exemplar. Utilizing the estimated mixture parameters, we then construct a compact slide representation that can be readily used for a wide range of downstream tasks. By performing an extensive evaluation of PANTHER on subtyping and survival tasks using 13 datasets, we show that 1) PANTHER outperforms or is on par with supervised MIL baselines and 2) the analysis of morphological prototypes brings new qualitative and quantitative insights into model interpretability.
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。
