Demystifying 3D Vision Transformers for Volumetric Brain MRI Tumor Detection
Standard 2D slice-by-slice neural networks often destroy cross-slice spatial continuum. In this article, we explain how spatial self-attention and 3D patch embeddings enable zero-shot tumor localization across unannotated volumetric scans.