[1]慕光睿,杨燕平,高耀宗,等.基于多尺度三维卷积神经网络的头颈部危及器官分割方法[J].南方医科大学学报,2020,(04):491-498.[doi:10.12122/j.issn.1673-4254.2020.04.07]
 Multi-scale 3D convolutional neural network-based segmentation of head and neck organs at risk[J].Journal of Southern Medical University,2020,(04):491-498.[doi:10.12122/j.issn.1673-4254.2020.04.07]
点击复制

基于多尺度三维卷积神经网络的头颈部危及器官分割方法()
分享到:

《南方医科大学学报》[ISSN:1673-4254/CN:44-1627/R]

卷:
期数:
2020年04期
页码:
491-498
栏目:
出版日期:
2020-04-30

文章信息/Info

Title:
Multi-scale 3D convolutional neural network-based segmentation of head and neck organs at risk
作者:
慕光睿杨燕平高耀宗冯前进
关键词:
器官分割卷积神经网络多尺度CT图像
Keywords:
organ segmentation convolutional neural network multi scales computed tomography
DOI:
10.12122/j.issn.1673-4254.2020.04.07
文献标志码:
A
摘要:
目的 研究一种基于三维卷积神经网络的CT图像头颈部危及器官分割算法。方法 本文构建了一个基于V-Net模型的头颈部危及器官自动分割算法。为了增强分割模型的特征表达能力,将SE(Squeeze-and-Excitation)模块与V-Net模型中残差卷 积模块相结合,提高与分割任务相关性更大的特征通道权重;采用多尺度策略,使用粗定位和精分割两个级联模型完成器官分割,其中输入图像在预处理时重采样为不同分辨率,使得模型分别专注于全局位置信息和局部细节特征的提取。结果 我们在头颈部22个危及器官的分割实验表明,相比于已有方法,本文提出的方法分割平均精度提升了9%,同时平均测试时间从33.82 s降低至2.79 s。结论 基于多尺度策略的三维卷积神经网络达到了更好的分割精度,且耗时极短,有望在临床应用中提高医生的工作效率。
Abstract:
Objective To establish an algorithm based on 3D convolution neural network to segment the organs at risk (OARs) in the head and neck on CT images. Methods We propose an automatic segmentation algorithm of head and neck OARs based on V-Net. To enhance the feature expression ability of the 3D neural network, we combined the squeeze and exception (SE) module with the residual convolution module in V-Net to increase the weight of the features that has greater contributions to the segmentation task. Using a multi-scale strategy, we completed organ segmentation using two cascade models for location and fine segmentation, and the input image was resampled to different resolutions during preprocessing to allow the two models to focus on the extraction of global location information and local detail features respectively. Results Our experiments on segmentation of 22 OARs in the head and neck indicated that compared with the existing methods, the proposed method achieved better segmentation accuracy and efficiency, and the average segmentation accuracy was improved by 9% . At the same time, the average test time was reduced from 33.82 s to 2.79 s. Conclusion The 3D convolution neural network based on multi-scale strategy can effectively and efficiently improve the accuracy of organ segmentation and can be potentially used in clinical setting for segmentation of other organs to improve the efficiency of clinical treatment.

相似文献/References:

[1]杜东阳,路利军,符瑞阳,等.手掌静脉识别:基于端到端卷积神经网络方法[J].南方医科大学学报,2019,(02):207.[doi:10.12122/j.issn.1673-4254.2019.02.13]
[2]邓力,傅蓉.基于心拍的端到端心律失常分类[J].南方医科大学学报,2019,(09):1071.[doi:10.12122/j.issn.1673-4254.2019.09.11]

更新日期/Last Update: 2020-04-30