Lung disease is one of the most common medical conditions in the world and it refers to several types of diseases or disorders that prevent the lungs from functioning properly. The main goal of the SoftLungX project is to create a ground-breaking, state-of-the-art software, completely adaptable to the possibilities of respiratory image-based disease recognition and risk level assessment. Bioengineering Research and Development Centre will be a technology provider to create a comprehensive deep learning model capable of recognizing and distinguishing between instances of COVID-19, emphysema, hernia, mass, pneumonia, pulmonary edema etc. using patients’ radiological scans and then classify these patients into multiple distinct risk classes based on the severity of the disease. While a lot of work suggests that neural networks may be used for the detection of respiratory disease, there seems to be a lack of systems that are capable of distinguishing COVID-19 from other lung structure and appearance altering diseases, hence, this project will be based not only on the identification of patients with COVID-19, but also on these previously mentioned diseases and the assessment of their risk levels. The main challenge in the project comes from the sheer amount of medical scans needed for every one of the aforementioned diseases that are required for the training of the deep learning model, and that all need to be labelled by their respective risk classes by a medical professional. This challenge is tackled by the use of data from public databases, combined with the new data that could be amassed in a timely manner for University Clinical Centre Kragujevac, which will also be the first technology adopter. Created system can be used in daily clinical practice, where it would assist pulmonologists and other medical experts in treatment of the patients suffering with respiratory illnesses.