OXnet: Omni-supervised Thoracic Disease Detection from Chest X-rays

TL;DR

OXnet is a novel deep learning framework that leverages multiple levels of supervision, including image-level and lesion-level annotations, to improve thoracic disease detection from chest X-rays, especially in data-scarce scenarios.

Contribution

This paper introduces the first omni-supervised detection network for chest X-rays, combining global classification, dual attention, prototype alignment, and pseudo-label distillation.

Findings

Outperforms existing methods on large-scale CXR dataset

Effectively utilizes various supervision granularities

Reduces annotation dependency for medical diagnosis

Abstract

Chest X-ray (CXR) is the most typical diagnostic X-ray examination for screening various thoracic diseases. Automatically localizing lesions from CXR is promising for alleviating radiologists' reading burden. However, CXR datasets are often with massive image-level annotations and scarce lesion-level annotations, and more often, without annotations. Thus far, unifying different supervision granularities to develop thoracic disease detection algorithms has not been comprehensively addressed. In this paper, we present OXnet, the first deep omni-supervised thoracic disease detection network to our best knowledge that uses as much available supervision as possible for CXR diagnosis. We first introduce supervised learning via a one-stage detection model. Then, we inject a global classification head to the detection model and propose dual attention alignment to guide the global gradient to the local detection branch, which enables learning lesion detection from image-level annotations. We also impose intra-class compactness and inter-class separability with global prototype alignment to further enhance the global information learning. Moreover, we leverage a soft focal loss to distill the soft pseudo-labels of unlabeled data generated by a teacher model. Extensive experiments on a large-scale chest X-ray dataset show the proposed OXnet outperforms competitive methods with significant margins. Further, we investigate omni-supervision under various annotation granularities and corroborate OXnet is a promising choice to mitigate the plight of annotation shortage for medical image diagnosis.