ADAptation: Reconstruction-based Unsupervised Active Learning for Breast Ultrasound Diagnosis
This work addresses the problem of costly data annotation and domain shifts in medical imaging for clinicians, offering an incremental improvement by combining diffusion models with active learning for domain adaptation.
The paper tackles performance drops in deep learning-based breast ultrasound diagnosis due to distribution shifts between training and test domains by proposing ADAptation, an unsupervised active learning framework that selects informative samples under limited annotation budgets, achieving superior results compared to existing AL-based methods across four datasets and five classifiers.
Deep learning-based diagnostic models often suffer performance drops due to distribution shifts between training (source) and test (target) domains. Collecting and labeling sufficient target domain data for model retraining represents an optimal solution, yet is limited by time and scarce resources. Active learning (AL) offers an efficient approach to reduce annotation costs while maintaining performance, but struggles to handle the challenge posed by distribution variations across different datasets. In this study, we propose a novel unsupervised Active learning framework for Domain Adaptation, named ADAptation, which efficiently selects informative samples from multi-domain data pools under limited annotation budget. As a fundamental step, our method first utilizes the distribution homogenization capabilities of diffusion models to bridge cross-dataset gaps by translating target images into source-domain style. We then introduce two key innovations: (a) a hypersphere-constrained contrastive learning network for compact feature clustering, and (b) a dual-scoring mechanism that quantifies and balances sample uncertainty and representativeness. Extensive experiments on four breast ultrasound datasets (three public and one in-house/multi-center) across five common deep classifiers demonstrate that our method surpasses existing strong AL-based competitors, validating its effectiveness and generalization for clinical domain adaptation. The code is available at the anonymized link: https://github.com/miccai25-966/ADAptation.