Summary:
- Fusion of UNET and Transformer.
- Deep learning applied to healthcare.
This project leverages deep learning to identify the tumor locations in MRI images. Recently, encoder-decoder architectures (e.g. U-net) have become the de facto standard for image segmentation. However, the locality of convolutional layers limits the global context and long-range spatial dependencies. These limitations are resolved through the fusion between Transformers, which extract long-range dependencies with self-attention mechanisms, with U-Net which captures low-level spatial details efficiently.
