arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2408.00427cs.CVcs.AI
收藏

CARMIL:面向全切片图像的多实例学习模型的上下文感知正则化

CARMIL: Context-Aware Regularization on Multiple Instance Learning models for Whole Slide Images

Owkin, Inc.(奥金公司);University of Oxford(牛津大学)

Thiziri Nait Saada, Valentina Di Proietto, Benoit Schmauch, Katharina Von Loga, Lucas Fidon

AI总结:

本文提出CARMIL正则化方案,以通用方式将空间上下文知识无缝集成到任意多实例学习模型中,并设计新的上下文感知度量指标,在TCGA GBM和TCGA COAD生存分析任务中验证其有效性。

摘要:

多实例学习(MIL)模型已被证明在基于全切片图像的癌症预后预测中有效。然而,原始的MIL形式错误地假设同一图像中的图像块彼此独立,导致信息在网络中传递时空间上下文丢失。鉴于癌细胞倾向于形成簇以及肿瘤存在空间指示特征,将上下文知识纳入预测尤为重要。当前最先进的方法通常使用注意力机制,并最终与图结合来捕获空间知识。在本文中,我们采用一种新颖且具有横向普适性的方法,从正则化的角度解决这一问题。我们提出了面向多实例学习的上下文感知正则化(CARMIL),这是一种通用的正则化方案,旨在将空间知识无缝集成到任何MIL模型中。此外,我们提出了一种新的通用度量指标,用于量化任何MIL模型应用于全切片图像时的上下文感知能力,填补了该领域此前未被探索的空白。我们在胶质母细胞瘤(TCGA GBM)和结肠癌数据(TCGA COAD)上的两个生存分析任务中评估了该框架的有效性。

英文摘要:

Multiple Instance Learning (MIL) models have proven effective for cancer prognosis from Whole Slide Images. However, the original MIL formulation incorrectly assumes the patches of the same image to be independent, leading to a loss of spatial context as information flows through the network. Incorporating contextual knowledge into predictions is particularly important given the inclination for cancerous cells to form clusters and the presence of spatial indicators for tumors. State-of-the-art methods often use attention mechanisms eventually combined with graphs to capture spatial knowledge. In this paper, we take a novel and transversal approach, addressing this issue through the lens of regularization. We propose Context-Aware Regularization for Multiple Instance Learning (CARMIL), a versatile regularization scheme designed to seamlessly integrate spatial knowledge into any MIL model. Additionally, we present a new and generic metric to quantify the Context-Awareness of any MIL model when applied to Whole Slide Images, resolving a previously unexplored gap in the field. The efficacy of our framework is evaluated for two survival analysis tasks on glioblastoma (TCGA GBM) and colon cancer data (TCGA COAD).

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。

↑