Liver CT image segmentation method based on CNN and Transformer(PDF)
《中国医学物理学杂志》[ISSN:1005-202X/CN:44-1351/R]
- Issue:
- 2023年第4期
- Page:
- 423-428
- Research Field:
- 医学影像物理
- Publishing date:
Info
- Title:
- Liver CT image segmentation method based on CNN and Transformer
- Author(s):
- HU Xiaoyang; LI Zhe
- School of Automation and Electrical Engineering, Shenyang Ligong University, Shenyang 110000, China
- Keywords:
- Keywords: convolutional neural network liver image segmentation multi-head self-attention mechanism atrous convolution
- PACS:
- R318；TP391
- DOI:
- DOI:10.3969/j.issn.1005-202X.2023.04.005
- Abstract:
- Abstract: Aiming at the problem of low accuracy of the existing convolutional neural network in liver image segmentation, a segmentation algorithm based on U-Net model is presented. The segmentation accuracy is improved using multi-head self-attention mechanism which was introduced into the skip connection of U-Net, atrous convolution in the encoder, and mixed loss function. The experimental results on LITS data set show that the Dice coefficient, mean itersection over union and mean pixel accuracy of liver segmentation using the proposed method are improved by 3.3%, 2.4% and 3.66% as compared with traditional U-Net method.
Last Update: 2023-04-25