Knowledge Distillation with Adaptive Asymmetric Label Sharpening for Semi-supervised Fracture Detection in Chest X-rays

TL;DR

This paper introduces a novel semi-supervised learning method with adaptive asymmetric label sharpening for improved fracture detection in chest X-rays, effectively utilizing large-scale image-level labels and limited expert annotations.

Contribution

The study proposes a new knowledge distillation approach with AALS to address label imbalance and leverage medical record data for better fracture detection in chest X-rays.

Findings

Achieved state-of-the-art AUROC of 0.9318 for rib fractures.

Improved FROC score to 0.8914, outperforming previous methods.

Demonstrated consistent performance gains across fracture types.

Abstract

Exploiting available medical records to train high performance computer-aided diagnosis (CAD) models via the semi-supervised learning (SSL) setting is emerging to tackle the prohibitively high labor costs involved in large-scale medical image annotations. Despite the extensive attentions received on SSL, previous methods failed to 1) account for the low disease prevalence in medical records and 2) utilize the image-level diagnosis indicated from the medical records. Both issues are unique to SSL for CAD models. In this work, we propose a new knowledge distillation method that effectively exploits large-scale image-level labels extracted from the medical records, augmented with limited expert annotated region-level labels, to train a rib and clavicle fracture CAD model for chest X-ray (CXR). Our method leverages the teacher-student model paradigm and features a novel adaptive asymmetric label sharpening (AALS) algorithm to address the label imbalance problem that specially exists in medical domain. Our approach is extensively evaluated on all CXR (N = 65,845) from the trauma registry of anonymous hospital over a period of 9 years (2008-2016), on the most common rib and clavicle fractures. The experiment results demonstrate that our method achieves the state-of-the-art fracture detection performance, i.e., an area under receiver operating characteristic curve (AUROC) of 0.9318 and a free-response receiver operating characteristic (FROC) score of 0.8914 on the rib fractures, significantly outperforming previous approaches by an AUROC gap of 1.63% and an FROC improvement by 3.74%. Consistent performance gains are also observed for clavicle fracture detection.