Department of Electrical, Information, Materials and Materials Engineering, Meijo University, Tempaku-ku, Nagoya, Aichi, 468-8502, Japan.
Department of Electrical and Electronic Engineering, Meijo University, Tempaku-ku, Nagoya, Aichi, Japan.
Int J Comput Assist Radiol Surg. 2024 Nov;19(11):2187-2193. doi: 10.1007/s11548-023-03038-2. Epub 2023 Dec 19.
Semantic segmentation of tubular structures, such as blood vessels and cell membranes, is a very difficult task, and it tends to break many predicted regions in the middle. This problem is due to the fact that tubular ground truth is very thin, and the number of pixels is extremely unbalanced compared to the background.
We present a novel training method using pseudo-labels generated by morphological transformation. Furthermore, we present an attention module using thickened pseudo-labels, called the expanded tube attention (ETA) module. By using the ETA module, the network learns thickened regions based on pseudo-labels at first and then gradually learns thinned original regions while transferring information in the thickened regions as an attention map.
Through experiments conducted on retina vessel image datasets using various evaluation measures, we confirmed that the proposed method using ETA modules improved the clDice metric accuracy in comparison with the conventional methods.
We demonstrated that the proposed novel expanded tube attention module using thickened pseudo-labels can achieve easy-to-hard learning.
管状结构（如血管和细胞膜）的语义分割是一项非常困难的任务，它往往会在中间断开许多预测区域。这个问题是由于管状真实情况非常细，与背景相比，像素数量极不平衡。
我们提出了一种使用形态变换生成伪标签的新训练方法。此外，我们提出了一个使用加厚伪标签的注意力模块，称为扩展管注意力（ETA）模块。通过使用 ETA 模块，网络首先基于伪标签学习加厚区域，然后在传输加厚区域中的信息作为注意力图的同时，逐渐学习变薄的原始区域。
通过使用各种评估指标在视网膜血管图像数据集上进行的实验，我们证实了与传统方法相比，使用 ETA 模块的建议方法提高了 clDice 度量精度。
我们证明了使用加厚伪标签的新型扩展管注意力模块可以实现从易到难的学习。
