Label Cleaning Multiple Instance Learning: Refining Coarse Annotations on Single Whole-Slide Images

TL;DR

This paper introduces LC-MIL, a novel method for refining coarse pathology annotations on single whole-slide images, improving accuracy without external data, and reducing pathologists' workload.

Contribution

The paper presents LC-MIL, a lightweight multiple instance learning approach that refines coarse annotations on individual WSIs without requiring large training datasets.

Findings

LC-MIL significantly improves annotation accuracy.

Outperforms state-of-the-art methods on diverse cancer datasets.

Effective in refining real pathologist annotations.

Abstract

Annotating cancerous regions in whole-slide images (WSIs) of pathology samples plays a critical role in clinical diagnosis, biomedical research, and machine learning algorithms development. However, generating exhaustive and accurate annotations is labor-intensive, challenging, and costly. Drawing only coarse and approximate annotations is a much easier task, less costly, and it alleviates pathologists' workload. In this paper, we study the problem of refining these approximate annotations in digital pathology to obtain more accurate ones. Some previous works have explored obtaining machine learning models from these inaccurate annotations, but few of them tackle the refinement problem where the mislabeled regions should be explicitly identified and corrected, and all of them require a -- often very large -- number of training samples. We present a method, named Label Cleaning Multiple Instance Learning (LC-MIL), to refine coarse annotations on a single WSI without the need of external training data. Patches cropped from a WSI with inaccurate labels are processed jointly within a multiple instance learning framework, mitigating their impact on the predictive model and refining the segmentation. Our experiments on a heterogeneous WSI set with breast cancer lymph node metastasis, liver cancer, and colorectal cancer samples show that LC-MIL significantly refines the coarse annotations, outperforming state-of-the-art alternatives, even while learning from a single slide. Moreover, we demonstrate how real annotations drawn by pathologists can be efficiently refined and improved by the proposed approach. All these results demonstrate that LC-MIL is a promising, light-weight tool to provide fine-grained annotations from coarsely annotated pathology sets.