Graduate School of Information Science and Technology, Hokkaido University, N-14, W-9, Kita-Ku, Sapporo, 060-0814, Japan.
Faculty of Information Science and Technology, Hokkaido University, N-14, W-9, Kita-Ku, Sapporo, 060-0814, Japan.
Comput Biol Med. 2023 May;158:106877. doi: 10.1016/j.compbiomed.2023.106877. Epub 2023 Mar 31.
Detecting COVID-19 from chest X-ray (CXR) images has become one of the fastest and easiest methods for detecting COVID-19. However, the existing methods usually use supervised transfer learning from natural images as a pretraining process. These methods do not consider the unique features of COVID-19 and the similar features between COVID-19 and other pneumonia.
In this paper, we want to design a novel high-accuracy COVID-19 detection method that uses CXR images, which can consider the unique features of COVID-19 and the similar features between COVID-19 and other pneumonia.
Our method consists of two phases. One is self-supervised learning-based pertaining; the other is batch knowledge ensembling-based fine-tuning. Self-supervised learning-based pretraining can learn distinguished representations from CXR images without manually annotated labels. On the other hand, batch knowledge ensembling-based fine-tuning can utilize category knowledge of images in a batch according to their visual feature similarities to improve detection performance. Unlike our previous implementation, we introduce batch knowledge ensembling into the fine-tuning phase, reducing the memory used in self-supervised learning and improving COVID-19 detection accuracy.
On two public COVID-19 CXR datasets, namely, a large dataset and an unbalanced dataset, our method exhibited promising COVID-19 detection performance. Our method maintains high detection accuracy even when annotated CXR training images are reduced significantly (e.g., using only 10% of the original dataset). In addition, our method is insensitive to changes in hyperparameters.
The proposed method outperforms other state-of-the-art COVID-19 detection methods in different settings. Our method can reduce the workloads of healthcare providers and radiologists.
从胸部 X 光（CXR）图像中检测 COVID-19 已成为检测 COVID-19 最快、最简单的方法之一。然而，现有的方法通常使用基于自然图像的监督迁移学习作为预训练过程。这些方法没有考虑到 COVID-19 的独特特征以及 COVID-19 和其他肺炎之间的相似特征。
本文旨在设计一种新颖的高精度 COVID-19 检测方法，该方法使用 CXR 图像，可以考虑 COVID-19 的独特特征以及 COVID-19 和其他肺炎之间的相似特征。
我们的方法由两个阶段组成。一个是基于自我监督学习的训练；另一个是基于批量知识集成的微调。基于自我监督学习的预训练可以从 CXR 图像中学习到有区别的表示，而无需手动标记标签。另一方面，基于批量知识集成的微调可以根据其视觉特征相似性利用批量图像的类别知识来提高检测性能。与我们之前的实现不同，我们在微调阶段引入了批量知识集成，减少了自我监督学习中使用的内存，并提高了 COVID-19 的检测准确性。
在两个公开的 COVID-19 CXR 数据集上，即一个大型数据集和一个不平衡数据集上，我们的方法表现出了有前景的 COVID-19 检测性能。即使减少了注释的 CXR 训练图像（例如，仅使用原始数据集的 10%），我们的方法也能保持较高的检测准确性。此外，我们的方法对超参数的变化不敏感。
提出的方法在不同的设置下优于其他最先进的 COVID-19 检测方法。我们的方法可以减轻医疗保健提供者和放射科医生的工作量。
