Collaborative learning methods for medical image segmentation are often variants of UNet, where the constructions of classifiers depend on each other and their outputs are supervised independently. However, they cannot explicitly ensure that optimizing auxiliary classifier heads leads to improved segmentation of target classifier. To resolve this problem, we propose a structured collaborative learning (SCL) method, which consists of a context-aware structured classifier population generation (CA-SCPG) module, where the feature propagation of the target classifier path is directly enhanced by the outputs of auxiliary classifiers via a light-weighted high-level context-aware dense connection (HLCA-DC) mechanism, and a knowledge-aware structured classifier population supervision (KA-SCPS) module, where the auxiliary classifiers are properly supervised under the guidance of target classifier's segmentations. Specifically, SCL is proposed based on a recurrent-dense-siamese decoder (RDS-Decoder), which consists of multiple siamese-decoder paths. CA-SCPG enhances the feature propagation of the decoder paths by HLCA-DC, which densely reuses previous decoder paths' output predictions to belong to the target classes as inputs to the latter decoder paths. KA-SCPS supervises the classifier heads simultaneously with KA-SCPS loss, which consists of a generalized weighted cross-entropy loss for deep class-imbalanced learning and a novel knowledge-aware Dice loss (KA-DL). KA-DL is a weighted Dice loss broadcasting knowledges learnt by the target classifier to other classifier heads, harmonizing the learning process of the classifier population. Experiments are performed based on PET/CT volumes with malignant melanoma, lymphoma, or lung cancer. Experimental results demonstrate the superiority of our SCL, when compared to the state-of-the-art methods and baselines.

中文翻译：

---

具有分布式在线学习的宽深图神经网络


用于医学图像分割的协作学习方法通​​常是 UNet 的变体，其中分类器的构造相互依赖，并且它们的输出受到独立监督。然而，他们无法明确确保优化辅助分类器头可以改善目标分类器的分割。为了解决这个问题，我们提出了一种结构化协作学习（SCL）方法，该方法由上下文感知结构化分类器群体生成（CA-SCPG）模块组成，其中目标分类器路径的特征传播直接由通过轻量级高级上下文感知密集连接（HLCA-DC）机制和知识感知结构化分类器群体监督（KA-SCPS）模块来辅助分类器，其中辅助分类器在目标的指导下受到适当的监督分类器的分段。具体来说，SCL是基于循环密集暹罗解码器（RDS-Decoder）提出的，该解码器由多个暹罗解码器路径组成。 CA-SCPG 通过 HLCA-DC 增强了解码器路径的特征传播，HLCA-DC 密集地重用先前解码器路径的输出预测以属于目标类别，作为后面解码器路径的输入。 KA-SCPS 与 KA-SCPS 损失同时监督分类器头，KA-SCPS 损失由用于深度类不平衡学习的广义加权交叉熵损失和新颖的知识感知 Dice 损失（KA-DL）组成。 KA-DL是一种加权Dice损失，将目标分类器学到的知识传播到其他分类器头，协调分类器群体的学习过程。实验是根据恶性黑色素瘤、淋巴瘤或肺癌的 PET/CT 体积进行的。 实验结果证明了我们的 SCL 与最先进的方法和基线相比的优越性。