UNETR: Transformers for 3D Medical Image Segmentation

Fully Convolutional Neural Networks (FCNNs) with contracting and expanding paths have shown prominence for the majority of medical image segmentation applications since the past decade. In FCNNs, the encoder plays an integral role by learning both global and local features and contextual representations which can be utilized for semantic output prediction by the decoder. Despite their success, the locality of convolutional layers in FCNNs, limits the capability of learning long-range spatial dependencies. Inspired by the recent success of transformers for Natural Language Processing (NLP) in long-range sequence learning, we reformulate the task of volumetric (3D) medical image segmentation as a sequence-to-sequence prediction problem. We introduce a novel architecture, dubbed as UNEt TRansformers (UNETR), that utilizes a transformer as the encoder to learn sequence representations of the input volume and effectively capture the global multi-scale information, while also following the successful U-shaped network design for the encoder and decoder. The transformer encoder is directly connected to a decoder via skip connections at different resolutions to compute the final semantic segmentation output. We have validated the performance of our method on the Multi Atlas Labeling Beyond The Cranial Vault (BTCV) dataset for multi-organ segmentation and the Medical Segmentation Decathlon (MSD) dataset for brain tumor and spleen segmentation tasks. Our benchmarks demonstrate new state-of-the-art peiformarce on the BTCV leaderboard.

Automated summary

Fully Convolutional Neural Networks (FCNNs) have been successful in medical image segmentation, but their limited ability to learn long-range dependencies is a challenge. Inspired by transformers in NLP, we propose a novel architecture called UNet Transformers (UNETR) to redefine volumetric medical image segmentation as a sequence prediction problem. By combining transformers and U-shaped network design in the encoder and decoder, we effectively capture global information and achieve semantic segmentation output.