[论文解读] GraphXCOVID: Explainable Deep Graph Diffusion Pseudo-Labelling for Identifying COVID-19 on Chest X-rays

Angelica I. Avilés-Rivero, Philip Sellars+2 more|arXiv (Cornell University)|Sep 30, 2020

COVID-19 diagnosis using AI参考文献 89被引用 34

一句话总结

GraphXCOVID 提出了一种基于图的深度半监督框架，通过扩散从少量有标签数据中生成伪标签来在胸部X光片上分类 COVID-19，在标签更少的情况下优于若干监督模型，并提供可解释的注意力图。

ABSTRACT

Can one learn to diagnose COVID-19 under extreme minimal supervision? Since the outbreak of the novel COVID-19 there has been a rush for developing Artificial Intelligence techniques for expert-level disease identification on Chest X-ray data. In particular, the use of deep supervised learning has become the go-to paradigm. However, the performance of such models is heavily dependent on the availability of a large and representative labelled dataset. The creation of which is a heavily expensive and time consuming task, and especially imposes a great challenge for a novel disease. Semi-supervised learning has shown the ability to match the incredible performance of supervised models whilst requiring a small fraction of the labelled examples. This makes the semi-supervised paradigm an attractive option for identifying COVID-19. In this work, we introduce a graph based deep semi-supervised framework for classifying COVID-19 from chest X-rays. Our framework introduces an optimisation model for graph diffusion that reinforces the natural relation among the tiny labelled set and the vast unlabelled data. We then connect the diffusion prediction output as pseudo-labels that are used in an iterative scheme in a deep net. We demonstrate, through our experiments, that our model is able to outperform the current leading supervised model with a tiny fraction of the labelled examples. Finally, we provide attention maps to accommodate the radiologist's mental model, better fitting their perceptual and cognitive abilities. These visualisation aims to assist the radiologist in judging whether the diagnostic is correct or not, and in consequence to accelerate the decision.

研究动机与目标

• 用尽可能少的标注数据实现对胸部X光片的鲁棒 COVID-19 分类。
• 开发一种基于图的扩散模型，为未标注数据生成伪标签。
• 将扩散产生的伪标签与深度网络结合，以迭代方式提升性能。
• 通过注意力图实现可解释性，帮助放射科医生诊断。

提出的方法

• 将数据集表示为一个图，其中节点是图像，边编码特征相似性。
• 通过深度网络提取特征；在学习到的特征空间内使用 k-NN 构建图。
• 在图上求解归一化且非光滑的 p=1 Dirichlet 能量扩散以获得伪标签。
• 使用加权交叉熵损失，考虑类别不平衡与不确定性，迭代更新伪标签和网络参数。
• 通过熵值计算伪标签的不确定性，以加权它们对训练的影响。

实验结果

研究问题

• RQ1 一个基于图的扩散模型是否能够生成可靠的伪标签，从而在有限标注的情况下实现深度半监督 COVID-19 分类？
• RQ2 将扩散生成的伪标签与深度网络结合，是否在 COVID-19 检测的准确性和灵敏度方面优于完全监督基线和现有 SSL 方法？
• RQ3 GraphXCOVID 生成的注意力图是否帮助放射科医生解读预测并提高诊断信心？

主要发现

• GraphXCOVID 在使用很小的标注数据情况下，对 COVID-19 具有较高的灵敏度且具有竞争力的准确率。
• 扩散基础的伪标签，结合不确定性加权和类别平衡，优于可比的半监督基线以及在 COVIDx 数据集上的若干监督模型。
• 该方法提供的注意力图与放射科医生的感知线索一致，支持诊断。