Abstract: Accurate and efficient segmentation of medical images is vital for clinical decisions but challenged by Vision Transformers’ lack of spatial precision and CNNs’ limited context. We propose ...