Li Weisheng, Qin Sheng, Li Feiyan, Wang Linhong
Chongqing Key Laboratory of Image Cognition, Chongqing University of Posts and Telecommunications, Chongqing, China.
Med Phys. 2021 Jan;48(1):329-341. doi: 10.1002/mp.14617. Epub 2020 Dec 7.
Pancreas segmentation is a difficult task because of the high intrapatient variability in the shape, size, and location of the organ, as well as the low contrast and small footprint of the CT scan. At present, the U-Net model is likely to lead to the problems of intraclass inconsistency and interclass indistinction in pancreas segmentation. To solve this problem, we improved the contextual and semantic feature information acquisition method of the biomedical image segmentation model (U-Net) based on a convolutional network and proposed an improved segmentation model called the multiscale attention dense residual U-shaped network (MAD-UNet).
There are two aspects considered in this method. First, we adopted dense residual blocks and weighted binary cross-entropy to enhance the semantic features to learn the details of the pancreas. Using such an approach can reduce the effects of intraclass inconsistency. Second, we used an attention mechanism and multiscale convolution to enrich the contextual information and suppress learning in unrelated areas. We let the model be more sensitive to pancreatic marginal information and reduced the impact of interclass indistinction.
We evaluated our model using fourfold cross-validation on 82 abdominal enhanced three-dimensional (3D) CT scans from the National Institutes of Health (NIH-82) and 281 3D CT scans from the 2018 MICCAI segmentation decathlon challenge (MSD). The experimental results showed that our method achieved state-of-the-art performance on the two pancreatic datasets. The mean Dice coefficients were 86.10% ± 3.52% and 88.50% ± 3.70%.
Our model can effectively solve the problems of intraclass inconsistency and interclass indistinction in the segmentation of the pancreas, and it has value in clinical application. Code is available at https://github.com/Mrqins/pancreas-segmentation.
胰腺分割是一项艰巨的任务，因为该器官在形状、大小和位置上存在较高的患者内部变异性，以及CT扫描的低对比度和小尺寸。目前，U-Net模型在胰腺分割中可能会导致类内不一致和类间区分不明显的问题。为了解决这个问题，我们改进了基于卷积网络的生物医学图像分割模型（U-Net）的上下文和语义特征信息获取方法，并提出了一种改进的分割模型，称为多尺度注意力密集残差U形网络（MAD-UNet）。
该方法考虑了两个方面。首先，我们采用密集残差块和加权二元交叉熵来增强语义特征，以学习胰腺的细节。使用这种方法可以减少类内不一致的影响。其次，我们使用注意力机制和多尺度卷积来丰富上下文信息，并抑制无关区域的学习。我们让模型对胰腺边缘信息更加敏感，并减少了类间区分不明显的影响。
我们使用四折交叉验证在来自美国国立卫生研究院（NIH-82）的82例腹部增强三维（3D）CT扫描和来自2018年医学图像计算与计算机辅助干预国际会议（MICCAI）分割十项全能挑战赛（MSD）的281例3D CT扫描上评估了我们的模型。实验结果表明，我们的方法在两个胰腺数据集上取得了领先的性能。平均Dice系数分别为86.10%±3.52%和88.50%±3.70%。
我们的模型可以有效解决胰腺分割中类内不一致和类间区分不明显的问题，并且在临床应用中具有价值。代码可在https://github.com/Mrqins/pancreas-segmentation获取。
