Weakly Supervised Tuberculosis Localization in Chest X-rays through Knowledge Distillation
This addresses the challenge of reducing spurious correlations and improving generalization for tuberculosis diagnosis in resource-limited settings, though it is incremental as it adapts an existing technique to a specific domain.
The study tackled the problem of localizing tuberculosis-related abnormalities in chest X-rays without needing bounding-box annotations by repurposing knowledge distillation, achieving a 0.2428 mIOU score and showing improved robustness over the teacher model.
Tuberculosis (TB) remains one of the leading causes of mortality worldwide, particularly in resource-limited countries. Chest X-ray (CXR) imaging serves as an accessible and cost-effective diagnostic tool but requires expert interpretation, which is often unavailable. Although machine learning models have shown high performance in TB classification, they often depend on spurious correlations and fail to generalize. Besides, building large datasets featuring high-quality annotations for medical images demands substantial resources and input from domain specialists, and typically involves several annotators reaching agreement, which results in enormous financial and logistical expenses. This study repurposes knowledge distillation technique to train CNN models reducing spurious correlations and localize TB-related abnormalities without requiring bounding-box annotations. By leveraging a teacher-student framework with ResNet50 architecture, the proposed method trained on TBX11k dataset achieve impressive 0.2428 mIOU score. Experimental results further reveal that the student model consistently outperforms the teacher, underscoring improved robustness and potential for broader clinical deployment in diverse settings.