A deep learning framework for pancreas segmentation with multi-atlas registration and 3D level-set
A deep learning framework for pancreas segmentation with multi-atlas registration and 3D level-set
复制标题
DOI：
10.1016/j.media.2020.101884
复制
发表时间：
2021-02-01
影响因子：
10.9
通讯作者：
Tang, Xiaoying
中科院分区：
文献类型：
--
作者：
Zhang, Yue;Wu, Jiong;Tang, Xiaoying
在本文中，我们提出并验证了一个深度学习框架，该框架结合了多图谱配准和水平集，用于从CT体积图像中分割胰腺。所提出的分割流水线由三个阶段组成，即粗、细和精阶段。首先，通过基于多图谱的三维几何配准和融合得到粗分割;之后，为了学习连接特征，基于3D补丁的卷积神经网络（CNN）和三个基于2D切片的CNN被联合用于基于从粗略分割确定的边界框来预测精细分割。最后，以精细分割为约束条件，采用三维水平集方法将原始图像信息与CNN概率图信息相结合，实现精细分割。换句话说，我们在所提出的框架中联合利用全局3D位置信息（注册），上下文信息（基于块的3D CNN），形状信息（基于切片的2.5D CNN）和边缘信息（3D水平集）。这些组件形成了我们的级联粗-精-精分割框架。我们在三个不同的数据集上测试了所提出的框架，这些数据集具有从不同资源获得的不同强度范围，分别包含36，82和281个CT体积图像。在每个数据集中，我们实现了超过82%的平均Dice评分，与其他现有的最先进的胰腺分割算法相比，具有上级优势或相当。(C)2020爱思唯尔B. V.保留所有权利。
In this paper, we propose and validate a deep learning framework that incorporates both multi-atlas registration and level-set for segmenting pancreas from CT volume images. The proposed segmentation pipeline consists of three stages, namely coarse, fine, and refine stages. Firstly, a coarse segmentation is obtained through multi-atlas based 3D diffeomorphic registration and fusion. After that, to learn the connection feature, a 3D patch-based convolutional neural network (CNN) and three 2D slice-based CNNs are jointly used to predict a fine segmentation based on a bounding box determined from the coarse segmentation. Finally, a 3D level-set method is used, with the fine segmentation being one of its constraints, to integrate information of the original image and the CNN-derived probability map to achieve a refine segmentation. In other words, we jointly utilize global 3D location information (registration), contextual information (patch-based 3D CNN), shape information (slice-based 2.5D CNN) and edge information (3D level-set) in the proposed framework. These components form our cascaded coarse-fine-refine segmentation framework. We test the proposed framework on three different datasets with varying intensity ranges obtained from different resources, respectively containing 36, 82 and 281 CT volume images. In each dataset, we achieve an average Dice score over 82%, being superior or comparable to other existing state-of-the-art pancreas segmentation algorithms. (C) 2020 Elsevier B.V. All rights reserved.