School of Mathematics and Statistics, Changsha University of Science and Technology, ChangSha 410114, China.
School of Mathematics and Statistics, Changsha University of Science and Technology, ChangSha 410114, China.
Comput Biol Med. 2024 Mar;170:108090. doi: 10.1016/j.compbiomed.2024.108090. Epub 2024 Feb 2.
The U-shaped convolutional neural network (CNN) has attained remarkable achievements in the segmentation of skin lesion. However, given the inherent locality of convolution, this architecture cannot capture long-range pixel dependencies and multiscale global contextual information effectively. Moreover, repeated convolutions and downsampling operations can readily result in the omission of intricate local fine-grained details. In this paper, we proposed a U-shaped network (DBNet-SI) equipped with a dual-branch module that combines shift window attention and inception structures. First, we proposed a dual-branch module that combines shift window attention and inception structures (MSI) to better capture multiscale global contextual information and long-range pixel dependencies. Specifically, we have devised a cross-branch bidirectional interaction module within the MSI module to enable information complementarity between the two branches in the channel and spatial dimensions. Therefore, MSI is capable of extracting distinguishing and comprehensive features to accurately identify the skin lesion boundaries. Second, we have devised a progressive feature enhancement and information compensation module (PFEIC), which progressively compensates for fine-grained features through reconstructed skip connections and integrated global context attention modules. The results of the experiment show the superior segmentation performance of DBNet-SI compared with other deep learning models for skin lesion segmentation in the ISIC2017 and ISIC2018 datasets. Ablation studies demonstrate that our model can effectively extract rich multiscale global contextual information and compensate for the loss of local details.
U 形卷积神经网络（CNN）在皮肤病变分割中取得了显著的成果。然而，由于卷积的固有局部性，这种架构无法有效地捕捉长程像素依赖关系和多尺度全局上下文信息。此外，重复的卷积和下采样操作很容易导致复杂的局部细粒度细节的丢失。在本文中，我们提出了一种 U 形网络（DBNet-SI），配备了一个双分支模块，该模块结合了移位窗口注意力和 inception 结构。首先，我们提出了一个双分支模块，该模块结合了移位窗口注意力和 inception 结构（MSI），以更好地捕捉多尺度全局上下文信息和长程像素依赖关系。具体来说，我们在 MSI 模块中设计了一个跨分支双向交互模块，以使两个分支在通道和空间维度上的信息具有互补性。因此，MSI 能够提取有区别和全面的特征，准确识别皮肤病变边界。其次，我们设计了一个渐进式特征增强和信息补偿模块（PFEIC），通过重建的跳过连接和集成的全局上下文注意力模块，逐步补偿细粒度特征。实验结果表明，DBNet-SI 在 ISIC2017 和 ISIC2018 数据集的皮肤病变分割方面，比其他深度学习模型具有更好的分割性能。消融研究表明，我们的模型可以有效地提取丰富的多尺度全局上下文信息，并补偿局部细节的丢失。
