Bilateral Asymmetry Guided Counterfactual Generating Network for Mammogram Classification

TL;DR

This paper introduces a novel bilateral asymmetry guided counterfactual network for mammogram classification, leveraging symmetric priors and causal modeling to improve lesion localization and classification accuracy without lesion annotations.

Contribution

It proposes a new counterfactual generation method based on bilateral symmetry and causal modeling, integrated into a generative adversarial network with a feedback mechanism.

Findings

Achieves state-of-the-art performance on INBreast dataset.

Effectively localizes lesions without explicit annotations.

Demonstrates the benefit of bilateral symmetry in mammogram analysis.

Abstract

Mammogram benign or malignant classification with only image-level labels is challenging due to the absence of lesion annotations. Motivated by the symmetric prior that the lesions on one side of breasts rarely appear in the corresponding areas on the other side, given a diseased image, we can explore a counterfactual problem that how would the features have behaved if there were no lesions in the image, so as to identify the lesion areas. We derive a new theoretical result for counterfactual generation based on the symmetric prior. By building a causal model that entails such a prior for bilateral images, we obtain two optimization goals for counterfactual generation, which can be accomplished via our newly proposed counterfactual generative network. Our proposed model is mainly composed of Generator Adversarial Network and a \emph{prediction feedback mechanism}, they are optimized jointly and prompt each other. Specifically, the former can further improve the classification performance by generating counterfactual features to calculate lesion areas. On the other hand, the latter helps counterfactual generation by the supervision of classification loss. The utility of our method and the effectiveness of each module in our model can be verified by state-of-the-art performance on INBreast and an in-house dataset and ablation studies.