Domain Adaptive Lung Nodule Detection in X-ray Image

TL;DR

This paper presents a novel domain adaptive method for lung nodule detection in X-ray images, combining mean teacher self-training, contrastive learning, and domain-invariant feature extraction to improve cross-center detection accuracy.

Contribution

The paper introduces a hierarchical contrastive learning strategy and a nodule-level domain-invariant feature learning module, along with a new dataset, to address domain shift in lung nodule detection.

Findings

Effective in reducing domain shift impacts

Improves detection accuracy across multiple datasets

Introduces a new annotated X-ray dataset

Abstract

Medical images from different healthcare centers exhibit varied data distributions, posing significant challenges for adapting lung nodule detection due to the domain shift between training and application phases. Traditional unsupervised domain adaptive detection methods often struggle with this shift, leading to suboptimal outcomes. To overcome these challenges, we introduce a novel domain adaptive approach for lung nodule detection that leverages mean teacher self-training and contrastive learning. First, we propose a hierarchical contrastive learning strategy to refine nodule representations and enhance the distinction between nodules and background. Second, we introduce a nodule-level domain-invariant feature learning (NDL) module to capture domain-invariant features through adversarial learning across different domains. Additionally, we propose a new annotated dataset of X-ray images to aid in advancing lung nodule detection research. Extensive experiments conducted on multiple X-ray datasets demonstrate the efficacy of our approach in mitigating domain shift impacts.