Computer-based analysis of early fibrosing lung disease (360G-Wellcome-209553_Z_17_Z)

Fibrosing lung disease (FLD) is an idiopathic condition, affecting older patients (median age=65 years) and smokers, accounting for 0.9% of all UK deaths in 2012. Unfortunately, despite newly available treatment options, FLD is typically diagnosed at an advanced stage with patients already markedly functionally impaired. Patient decline is often rapid. A constraint with diagnosing disease at an early stage is that subtle minor CT abnormalities that may evolve into rapidly progressive disease, are hard to identify and yet to be characterised visually. Large population studies may identify those subtle CT features portending progressive disease. Such large-scale analysis of CT imaging would be best suited to advanced computer analytic tools. We therefore aim to develop sophisticated computer tools to evaluate CTs in 20,000 heavy-smoker patients undergoing repeated chest imaging, as part of a lung cancer screening study, to identify early and potentially progressive FLD on CT. GOALS 1.Characterise baseline CT patterns indicative of early FLD and progressive FLD. 2. Predict the trajectory of a patient’s fibrosis progression using computer quantitation of change in an individuals CT features. 3. Generate population-wide quantitative CT metrics (age, gender and race specific) as a reference range applicable to other worldwide lung cancer screening studies.