ThreeF-Net: Fine-grained feature fusion network for breast ultrasound image segmentation.

Authors

Bian X,Liu J,Xu S,Liu W,Mei L,Xiao C,Yang F

Affiliations (6)

  • School Of Information Engineering, Yancheng Institute Of Technology, Hope Avenue Middle Road, Yancheng, 224051, Jiangsu, China. Electronic address: [email protected].
  • School Of Information Engineering, Yancheng Institute Of Technology, Hope Avenue Middle Road, Yancheng, 224051, Jiangsu, China.
  • College of Computer Engineering, Jimei University, Yinjiang Road No. 183, Xiamen, 361005, Fujian, China.
  • National Research Facility for Phenotypic & Genetic Analysis of Model Animals (Primate Facility), Key Laboratory of Genetic Evolution & Animal Models, and National Resource Center for Nonhuman Primates, Kunming Institute of Zoology, Chinese Academy of Sciences, East Jiaochang Road No. 32, Kunming, 650221, Yunnan, China.
  • National Key Laboratory of Radar Signal Processing, Xidian university, No. 2 South Taibai Road, Xian, 710071, Shaanxi, China.
  • School of Resources and Environmental Engineering, Wuhan University of Science and Technology, 947 Heping Avenue, Qingshan District, Xian, 430081, Wuhan, China.

Abstract

Convolutional Neural Networks (CNNs) have achieved remarkable success in breast ultrasound image segmentation, but they still face several challenges when dealing with breast lesions. Due to the limitations of CNNs in modeling long-range dependencies, they often perform poorly in handling issues such as similar intensity distributions, irregular lesion shapes, and blurry boundaries, leading to low segmentation accuracy. To address these issues, we propose the ThreeF-Net, a fine-grained feature fusion network. This network combines the advantages of CNNs and Transformers, aiming to simultaneously capture local features and model long-range dependencies, thereby improving the accuracy and stability of segmentation tasks. Specifically, we designed a Transformer-assisted Dual Encoder Architecture (TDE), which integrates convolutional modules and self-attention modules to achieve collaborative learning of local and global features. Additionally, we designed a Global Group Feature Extraction (GGFE) module, which effectively fuses the features learned by CNNs and Transformers, enhancing feature representation ability. To further improve model performance, we also introduced a Dynamic Fine-grained Convolution (DFC) module, which significantly improves lesion boundary segmentation accuracy by dynamically adjusting convolution kernels and capturing multi-scale features. Comparative experiments with state-of-the-art segmentation methods on three public breast ultrasound datasets demonstrate that ThreeF-Net outperforms existing methods across multiple key evaluation metrics.

Topics

Journal Article

Ready to Sharpen Your Edge?

Join hundreds of your peers who rely on RadAI Slice. Get the essential weekly briefing that empowers you to navigate the future of radiology.

We respect your privacy. Unsubscribe at any time.