The process of generating medical reports from medical images using deep learning techniques is a rapidly growing field. This task is critical and time-consuming for radiologists. Previous approaches focused on producing text that is readable for humans but may not represent the clinical facts well. Despite the difficulty of creating medical descriptions from images, deep learning algorithms show promising results. The earlier research primarily concentrated on generating lengthy descriptions of images but ignored the characteristics of the image and the important patient background information that can aid in diagnosis. To address these challenges, a creative encoder-decoder paradigm was proposed using an attention mechanism focused on chest X-ray images and reports. The attention mechanism is capable of simultaneously concentrating on the spatial and channel details of an image and linking it to the corresponding sentence topic. The IU X-ray dataset was used for study, which was collected from Indiana University. The proposed framework predicts medical descriptions based on chest X-ray images and creates a comprehensible radiology report. The suggested method learns how to extract significant image features from the deep convolutional neural network (CNN) model and generates textual descriptions. Our experiments have shown that the proposed model outperforms all other baseline models.