HAU-Net: Hybrid CNN-transformer for breast ultrasound image segmentation

Huaikun Zhang, Jing Lian, Zetong Yi, Ruichao Wu, Xiangyu Lu, Pei Ma, Yide Ma

Abstract

Highlights•HAU-Net: Transformer-CNN fusion for accurate breast segmentation in ultrasound images.•L-G Transformer Block: long-range dependency modeling, reducing computational cost.•CAB: Enhanced feature interaction, higher segmentation accuracy.•Outperforms state-of-the-art: Higher Dice, IoU, Accuracy, and more on public datasets.•Robust segmentation: Advantageous for malignant, large-sized tumors segmentation.
BibTeX
@article{zhang2024hau,
  author = {Huaikun Zhang and Jing Lian and Zetong Yi and Ruichao Wu and Xiangyu Lu and Pei Ma and Yide Ma},
  title = {{HAU-Net: Hybrid CNN-transformer for breast ultrasound image segmentation}},
  journal = {Biomedical Signal Processing and Control},
  year = {2024},
  url = {https://openreview.net/forum?id=9sfuwGT1EB},
}

Loading related content...

Figure