华南理工大学学报(自然科学版) ›› 2025, Vol. 53 ›› Issue (5): 94-108.doi: 10.12141/j.issn.1000-565X.240439

• 计算机科学与技术 • 上一篇    

多尺度特征融合与重构卷积的肝肿瘤分割方法

 马金林1,2　酒志青2†　马自萍3　夏明格2　张凯2　程叶霞4　马瑞士2   

  1. 1．北方民族大学 图形图像智能信息处理国家民委重点实验室，宁夏 银川 750021；

    2．北方民族大学 计算机科学与工程学院，宁夏 银川 750021；

    3．北方民族大学 数学与信息科学学院，宁夏 银川 750021；

    4．中国移动通信有限公司，北京 100033

  • 出版日期:2025-05-25 发布日期:2024-12-02

A Liver Tumor Segmentation Method Based on Multi-Scale Feature Fusion and Reconstruction Convolution

MA Jinlin1,2　JIU Zhiqing2　MA Ziping3　XIA Mingge2　ZHANG Kai2  CHENG Yexia4　MA Ruishi2   

  1. 1. Key Laboratory of Intelligent Information Processing of Image and Graphics，North Minzu University，Yinchuan 750021，Ningxia，China；

    2.School of Computer Science and Engineering，North Minzu University，Yinchuan 750021，Ningxia，China；

    3.School of Mathematics and Information Science，North Minzu University，Yinchuan 750021，Ningxia，China；

    4.China Mobile Communications Corporation，Beijing 100033，China

  • Online:2025-05-25 Published:2024-12-02

摘要：

针对肝肿瘤特征表达能力不足和全局上下文信息传递受限的问题，提出一种改进UNet的肝肿瘤分割方法。首先，设计一种低秩重构卷积，优化传统卷积运算所导致的大量参数问题，并用其构建使用残差结构改进编解码器的卷积核重构模块，使编码器保留更多的细节信息，使解码器更有效地恢复信息，以提升肝肿瘤特征的表达能力。然后，为丰富全局上下文信息的传递，设计三分支空间金字塔池化模块，优化瓶颈结构的信息传递，打破单一路径的限制。其次，设计多尺度特征融合模块，优化编码器信息的复用机制，增强模型对全局上下文信息的建模能力，并提升其在提取不同尺度肝肿瘤特征方面的效能。最后，在LiTS2017和3DIRCADb数据集上测试本文方法的性能。实验结果表明，本文方法在LiTS2017数据集肝脏分割任务中的Dice和IoU值达97.56%和95.25%，肝肿瘤分割任务中的Dice和IoU值达89.71%和81.58%。在3DIRCADb数据集肝脏分割任务中的Dice和IoU值达97.63%和95.39%，肝肿瘤分割任务中的Dice和IoU值达89.62%和81.63%。本文方法可以有效缓解肝肿瘤特征表达能力不足的问题，且进一步提升了模型捕捉全局上下文信息的能力。

关键词: 肝肿瘤分割, 卷积核重构, 空间金字塔池化, 多尺度特征融合

Abstract:

Aiming at the problem of insufficient expression ability of liver tumor features and limited global contextual information transmission, an improved UNet liver tumor segmentation method is proposed. Firstly, a low rank reconstruction convolution is designed to optimize the large number of parameter problems caused by traditional convolution operations, and use it to construct a convolution kernel reconstruction module that uses residual structure to improve the encoder decoder, so that the encoder retains more detailed information and the decoder recovers information more effectively, thereby enhancing the expression ability of liver tumor features. Then, to enrich the transmission of global contextual information, a three branch spatial pyramid pooling module is designed to optimize the bottleneck structure of information transmission and break the limitation of a single path. Secondly, a multi-scale feature fusion module is designed to optimize the reuse mechanism of encoder information, enhance the modeling ability of the model for global contextual information, and improve its efficiency in extracting liver tumor features at different scales. Finally, the performance of our method was tested on the LiTS2017 and 3DIRCADb datasets. The experimental results show that our method achieves Dice and IoU values of 97.56% and 95.25% in the liver segmentation task on the LiTS2017 dataset, and 89.71% and 81.58% in the liver tumor segmentation task, respectively. The Dice and IoU values in the liver segmentation task of the 3DIRCADb dataset reached 97.63% and 95.39%, respectively, while the Dice and IoU values in the liver tumor segmentation task reached 89.62% and 81.63%, respectively. This method can effectively alleviate the problem of insufficient expression ability of liver tumor features, and further enhance the model's ability to capture global contextual information.

Key words: liver tumor segmentation, convolutional kernel reconstruction, spatial pyramid pooling, multi scale feature fusion

总访问量
今日访问
在线人数
粤ICP备05084312号
版权所有 © 华南理工大学学报编辑部
地址：广州 五山 华南理工大学17号楼 　邮编：510640
电 话：020-87111794　E-mail : journal@scut.edu.cn
  • 微信号：Journalscut
  • QQ群：699235083