Lung cancer, often referred to as chest cancer, is predominantly attributed to exposure to carcinogens, which are substances known to cause cancer. The primary factor contributing to the development of lung cancer is cigarette smoking, encompassing both direct smoking and exposure to secondhand smoke. Chest cancer detection is very much desired at the earliest stage for better diagnosis. To enhance detection accuracy and achieve precise results, it is imperative to employ high-resolution inputs. Visual Super Resolution Network (VSRN) model specifically designed for image super-resolution of human lung images is proposed in this paper. Our model incorporates two key components: A Channel Spatial Attention Module and the Layer Feature Graph Representation Learning Module (LGRL). The LGRL module in our model plays a crucial role in capturing the interdependence among features derived from different layers. This module enables us to extract more fine-grained details from the images, enhancing the overall representation. Furthermore, we have integrated a Spatial Attention (SA) block and a Channel Attention Block (CAB). This integration allows us to consider both channel and spatial information of the image. By leveraging these attention modules, it allows us to improve the quality of the image. These obtained high-resolution images are used to diagnose chest cancer.