The paper introduces CSWin-UNet, a novel U-shaped segmentation method that integrates the CSWin self-attention mechanism into the UNet architecture. This approach enhances both computational efficiency and receptive field interactions, while also improving segmentation accuracy. The CSWin self-attention mechanism allows for horizontal and vertical stripe self-attention, broadening the focus area for each token and facilitating more comprehensive analysis. The decoder employs a CARAFE (Content-Aware ReAssembly of FEatures) layer for upsampling, which provides more accurate pixel-level segmentation masks. Experimental results on diverse datasets, including synapse multi-organ CT, cardiac MRI, and skin lesions, demonstrate that CSWin-UNet maintains low model complexity while delivering high segmentation accuracy. The method outperforms existing methods in terms of both computational efficiency and segmentation accuracy, making it a significant advancement in medical image segmentation.The paper introduces CSWin-UNet, a novel U-shaped segmentation method that integrates the CSWin self-attention mechanism into the UNet architecture. This approach enhances both computational efficiency and receptive field interactions, while also improving segmentation accuracy. The CSWin self-attention mechanism allows for horizontal and vertical stripe self-attention, broadening the focus area for each token and facilitating more comprehensive analysis. The decoder employs a CARAFE (Content-Aware ReAssembly of FEatures) layer for upsampling, which provides more accurate pixel-level segmentation masks. Experimental results on diverse datasets, including synapse multi-organ CT, cardiac MRI, and skin lesions, demonstrate that CSWin-UNet maintains low model complexity while delivering high segmentation accuracy. The method outperforms existing methods in terms of both computational efficiency and segmentation accuracy, making it a significant advancement in medical image segmentation.