PULMONARY FIBROSIS PREDICTION

Imagine that at some point, you started breathing shallowly and with constant effort. It took several months for you to be officially diagnosed with pulmonary fibrosis, a lung scarring-related condition for which there is no recognized cause or treatment. You would want to be aware of your prognosis if something happened to you. This is where a concerning illness can become overwhelming for the patient: results might vary from stable long-term conditions to rapidly declining health, and it can be difficult for physicians to predict a patient’s potential place on this spectrum. Even with access to a chest CT scan, current treatments for fibrotic lung illnesses are challenging. Furthermore, there are problems planning clinical studies due to the wide range of possible outcomes. Ultimately, individuals experience severe anxiety due to the illness’s opaque nature, in addition to symptoms associated with fibrosis. With the help of this initiative, we hope to forecast how much a patient’s lung function will deteriorate in response to a lung CT scan. We will use a spirometer to measure the amount of air that is inhaled and expelled in order to evaluate how well lung function is sustained. Using the image, metadata, and baseline FVC as input, the goal is to create a prediction using machine learning algorithms. If it is successful, patients with this incurable lung illness and their families will have a better understanding of their prognosis when they receive their initial diagnosis. Enhancing severity identification might also hasten the clinical development of new medicines and have a positive effect on treatment trial design.