COIN: Counterfactual inpainting for weakly supervised semantic segmentation for medical images

TL;DR

COIN is a novel counterfactual inpainting method that leverages weak supervision and explainable AI to accurately segment tumors in medical images without requiring detailed annotations.

Contribution

The paper introduces COIN, a new counterfactual inpainting approach that uses image-level labels and generative models to improve medical image segmentation.

Findings

COIN outperforms attribution methods like RISE, ScoreCAM, and LayerCAM.

COIN effectively segments synthetic and real kidney tumors from CT images.

The approach reduces reliance on extensive annotated datasets in medical imaging.

Abstract

Deep learning is dramatically transforming the field of medical imaging and radiology, enabling the identification of pathologies in medical images, including computed tomography (CT) and X-ray scans. However, the performance of deep learning models, particularly in segmentation tasks, is often limited by the need for extensive annotated datasets. To address this challenge, the capabilities of weakly supervised semantic segmentation are explored through the lens of Explainable AI and the generation of counterfactual explanations. The scope of this research is development of a novel counterfactual inpainting approach (COIN) that flips the predicted classification label from abnormal to normal by using a generative model. For instance, if the classifier deems an input medical image X as abnormal, indicating the presence of a pathology, the generative model aims to inpaint the abnormal region, thus reversing the classifier's original prediction label. The approach enables us to produce precise segmentations for pathologies without depending on pre-existing segmentation masks. Crucially, image-level labels are utilized, which are substantially easier to acquire than creating detailed segmentation masks. The effectiveness of the method is demonstrated by segmenting synthetic targets and actual kidney tumors from CT images acquired from Tartu University Hospital in Estonia. The findings indicate that COIN greatly surpasses established attribution methods, such as RISE, ScoreCAM, and LayerCAM, as well as an alternative counterfactual explanation method introduced by Singla et al. This evidence suggests that COIN is a promising approach for semantic segmentation of tumors in CT images, and presents a step forward in making deep learning applications more accessible and effective in healthcare, where annotated data is scarce.