The School of Future Technology, Tianjin University, Tianjin, 300072, China.
The School of Electrical and Information Engineering, Tianjin University, Tianjin, 300072, China.
Comput Biol Med. 2023 Oct;165:107398. doi: 10.1016/j.compbiomed.2023.107398. Epub 2023 Sep 9.
Semi-supervised learning aims to train a high-performance model with a minority of labeled data and a majority of unlabeled data. Existing methods mostly adopt the mechanism of prediction task to obtain precise segmentation maps with the constraints of consistency or pseudo-labels, whereas the mechanism usually fails to overcome confirmation bias. To address this issue, in this paper, we propose a novel Confidence-Guided Mask Learning (CGML) for semi-supervised medical image segmentation. Specifically, on the basis of the prediction task, we further introduce an auxiliary generation task with mask learning, which intends to reconstruct the masked images for extremely facilitating the model capability of learning feature representations. Moreover, a confidence-guided masking strategy is developed to enhance model discrimination in uncertain regions. Besides, we introduce a triple-consistency loss to enforce a consistent prediction of the masked unlabeled image, original unlabeled image and reconstructed unlabeled image for generating more reliable results. Extensive experiments on two datasets demonstrate that our proposed method achieves remarkable performance.
半监督学习旨在利用少数带标签的数据和多数无标签的数据训练高性能模型。现有的方法大多采用预测任务的机制，通过一致性或伪标签的约束来获得精确的分割图，而这种机制通常无法克服确认偏差。针对这个问题，本文提出了一种新颖的用于半监督医学图像分割的置信度引导掩模学习（CGML）方法。具体来说，在预测任务的基础上，我们进一步引入了一个带有掩模学习的辅助生成任务，旨在重建掩模图像，这极大地促进了模型学习特征表示的能力。此外，还提出了一种置信度引导掩蔽策略，以增强模型在不确定区域的判别能力。此外，还引入了三重一致性损失，以强制对掩蔽的未标记图像、原始未标记图像和重建的未标记图像进行一致的预测，从而生成更可靠的结果。在两个数据集上的广泛实验表明，所提出的方法取得了显著的性能。
