MultiTrans: Multi-branch transformer network for medical image segmentation.

Journal: Computer methods and programs in biomedicine
Published Date:

Abstract

BACKGROUND AND OBJECTIVE: Transformer, which is notable for its ability of global context modeling, has been used to remedy the shortcomings of Convolutional neural networks (CNN) and break its dominance in medical image segmentation. However, the self-attention module is both memory and computational inefficient, so many methods have to build their Transformer branch upon largely downsampled feature maps or adopt the tokenized image patches to fit their model into accessible GPUs. This patch-wise operation restricts the network in extracting pixel-level intrinsic structural or dependencies inside each patch, hurting the performance of pixel-level classification tasks.

Authors

  • Yanhua Zhang
    Department of Electronics and Telecommunications, Politecnico di Torino, Corso Duca degli Abruzzi 24, Turin, 10129, Italy; School of Astronautics, Northwestern Polytechnical University, 127 West Youyi Road, Xi'an, 710072, China. Electronic address: yanhua.zhang@studenti.polito.it.
  • Gabriella Balestra
  • Ke Zhang
    Center for Radiation Oncology, Affiliated Hangzhou Cancer Hospital, Zhejiang University School of Medicine, Hangzhou 310001, China.
  • Jingyu Wang
    Center of Medical & Health Analysis, School of Public Health, Peking University, Beijing, China.
  • Samanta Rosati
  • Valentina Giannini