CARMIL: Context-Aware Regularization on Multiple Instance Learning models for Whole Slide Images
This work addresses the need for better spatial context modeling in cancer prognosis from medical images, offering a versatile solution that can be applied to any MIL model, though it is incremental as it builds on existing regularization and MIL frameworks.
The authors tackled the problem of spatial context loss in Multiple Instance Learning models for Whole Slide Images by proposing CARMIL, a regularization scheme that integrates spatial knowledge, and introduced a new metric to quantify context-awareness, achieving competitive performance on glioblastoma and colon cancer survival analysis tasks.
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).