Multi-scale Hierarchical Vision Transformer with Cascaded Attention Decoding for Medical Image Segmentation
This work addresses generalization issues in medical image segmentation, which is crucial for healthcare applications, but it appears incremental as it builds on existing transformer methods.
The paper tackles the limited generalization of transformers in medical image segmentation by introducing a multi-scale hierarchical vision transformer (MERIT) and a cascaded attention decoder (CASCADE), achieving superior performance on benchmarks like Synapse Multi-organ and ACDC.
Transformers have shown great success in medical image segmentation. However, transformers may exhibit a limited generalization ability due to the underlying single-scale self-attention (SA) mechanism. In this paper, we address this issue by introducing a Multi-scale hiERarchical vIsion Transformer (MERIT) backbone network, which improves the generalizability of the model by computing SA at multiple scales. We also incorporate an attention-based decoder, namely Cascaded Attention Decoding (CASCADE), for further refinement of multi-stage features generated by MERIT. Finally, we introduce an effective multi-stage feature mixing loss aggregation (MUTATION) method for better model training via implicit ensembling. Our experiments on two widely used medical image segmentation benchmarks (i.e., Synapse Multi-organ, ACDC) demonstrate the superior performance of MERIT over state-of-the-art methods. Our MERIT architecture and MUTATION loss aggregation can be used with downstream medical image and semantic segmentation tasks.