ViTResUNet-Based-Medical-Image-Segmentation

ViTResUNet is a hybrid CNN-Transformer architecture with the backbone of TransUNet. The experiments were conducted on Synapse multi-organ segmentation dataset.

• Access to the synapse multi-organ dataset:

  • Sign up in the official https://www.synapse.org/ and download the dataset. Convert them to numpy format, clip the images within [-125, 275], normalize each 3D image to [0, 1], and extract 2D slices from 3D volume for training cases while keeping the 3D volume in h5 format for testing cases.
  • It is possible to request the preprocessed dataset from the original repo authors.

• The architecure is given below for better understanding

Vision Transformer ( You can watch this repo https://github.com/RadeenXALNW/Vision-Transformer-ViT ) for better understanding the intuition

Refereneces

• https://arxiv.org/pdf/2102.04306.pdf
• https://github.com/Beckschen/TransUNet
• https://arxiv.org/abs/1807.06521
• https://github.com/xmu-xiaoma666/External-Attention-pytorch