Segtransvae: Hybrid Cnn - Transformer With Regularization For Medical Image Segmentation

Current researches on deep learning for medical image segmentation expose their limitation in learning global semantic information and generalizing badly when trained on insufficient amounts of data. To tackle these issues, a novel network named SegTransVAE is proposed in this paper. SegTransVAE is built upon encoder-decoder architecture exploting Transformer with the variational autoencoder (VAE) branch to the network to reconstruct the input images jointly with segmentation. To the best of our knowledge, this is the first method combining the success of CNN, Transformer and VAE. Evaluation on a variety of recently introduced datasets shows that SegTransVAE outperforms previous methods in Dice Score and 95%-Haudorff Distance while having comparable inference time to simple CNN-based architecture network.