The insufficient resolution of medical images, especially the low spatial resolution in the depth direction, may lead to the loss of critical information, thereby affecting the accuracy of medical diagnosis. Super-resolution (SR) technology plays a crucial role in medical imaging by enhancing image resolution to provide more detailed structural information. However, traditional single-image super-resolution (SISR) methods struggle to fully exploit 3D spatial information, resulting in insufficient spatial consistency between slices, which leads to artifacts and discontinuous textures, limiting their applicability in 3D medical image reconstruction. To address these challenges, this paper proposes the N-gram Swin Transformer Network (NGSWN) for super-resolution of CT images, specifically aiming to address the issue of insufficient resolution in the depth direction. The proposed model adopts an asymmetric encoder-decoder structure and integrates an N-gram-based mechanism to enhance feature extraction and reconstruction capabilities. By leveraging spatial relationships between slices, the NGSWN generates high-resolution CT images with better continuity and fewer artifacts. Experimental results demonstrate that the NGSWN outperforms both traditional and state-of-the-art methods in terms of PSNR and SSIM metrics, highlighting its significant potential for enhancing medical imaging quality and improving diagnostic accuracy.