Login / Signup

Advantages of transformer and its application for medical image segmentation: a survey.

Qiumei PuZuoxin XiShuai YinZhe ZhaoLina Zhao
Published in: Biomedical engineering online (2024)
Even if there is a pure transformer model without any convolution operator, the sample size of medical picture segmentation still restricts the growth of the transformer, even though it can be relieved by a pretraining model. More often than not, researchers are still designing models using transformer and convolution operators.
Keyphrases
  • deep learning
  • healthcare
  • convolutional neural network
  • neural network