Lung diseases cause millions of deaths per year worldwide. For medical diagnosis, different medical images are available, with computed tomography (CT) standing out as effective tools for pathology detection. Recently, the use of artificial intelligence (AI) is an advance in early diagnosis, but the quality of the images in the acquisition greatly limits the result. Therefore, this paper proposes CTextureFusion, a new approach to improve the resolution of lung CT images by combining advanced super-resolution imaging techniques like the existing multi-modal and multi-head attention mechanisms and integrating low-resolution images with reference images and applying specific filters for edge detection and contrast enhancement, our model achieves high-quality, detailed image reconstruction. In experiments, the model demonstrates significant improvements in quantitative and qualitative terms. These results suggest a great potential of the model as digital preprocessing for further diagnostic enhancement.