关键词: 肾脏三维结构分割, 注意力模块, 计算机断层血管造影, 损失函数, 边缘检测

Abstract: To address the problems of accurate segmentation difficulties caused by large differences between different structures, small size of arteries and veins, thin structures and uneven grayscale distribution and artifacts in computed tomography angiography(CTA) images in kidney structures, a kidney 3D structure segmentation model MDAUnet (MultiDecoder-Attention-Unet) based on multi-decoder and attention mechanism with CTA was proposed. Firstly, to address the problem that the network cannot share weights due to large differences between different structures, a multi-decoder structure was used to match different decoder branches for feature structures with different semantic structures; secondly, to address the problem that it was difficult to segment blood vessels with small size and thin structures, an asymmetric spatial channel joint attention module was introduced to make the model more focused on tubular structures, and the learned feature information was simultaneously spatially dimensional and channel Finally, in order to ensure that the model paid enough attention to the vessel structure in backpropagation, an improved WHRA(weighted hard region adaptation) loss was proposed as a loss function to dynamically maintain the inter-class balance of the vessel structure during training as well as to preserve the characteristics of the background information; In addition, in order to improve the contrast of the grayscale values of the feature map, the traditional image processing edge detection operator was embedded into the pre-processing stage of the model, and the feature enhancement of the boundary of the region of interest to be segmented made the model paid more attention to the boundary information and suppressed the artifact information. The results show that the Dice similarity coefficient (DSC), hausdorff distance 95(HD95) and average surface distance (AVD) of the proposed MDAUnet model on the kidney structure segmentation task are 89.1%, 1.76mm and 1.04mm, respectively. Compared with MGANet, MDAUnet improves the DSC index by 1.2 percentage points; compared with UNETR, MDAUnet reduces HD95 and ASD indexes by 0.84 mm and 0.45 mm, respectively. It can be seen that MDAUnet can effectively improve the segmentation accuracy of the three-dimensional structure of the kidney, and help doctors to evaluate the condition objectively and effectively in clinical operations.

Key words: kidney Three-Dimensional(3D) structural segmentation, attention module, computed tomography angiography (CTA), loss function, edge detection