Interstitial lung diseases affect over 70,000 people in the UK. These diseases cause shortness of breath, coughing and tiredness. They also cause scarring in the lungs, which may get worse over time and can be fatal.
Lung scarring is known as pulmonary fibrosis. It progresses differently in individual patients and predicting who is likely to get worse is challenging. All patients with pulmonary fibrosis routinely undergo high resolution computed tomography scans (HRCT). HRCT produces detailed images of the lungs.
Radiologists review these scans to understand the extent and type of disease a patient has developed. However, individual radiologists can interpret scan results differently. This can limit the usefulness of HRCT as a tool to monitor disease and predict how it will progress.
Project aims
Our study will investigate whether artificial intelligence can more accurately predict how an individual patient’s disease might progress.
We will look at historic scan results for patients cared for by three hospitals in the South West. These patients will also have undergone lung function and exercise testing.
We will use AI to analyse the scan results. We will compare the AI results with how patients’ lung function and exercise capacity changed over time.
What we hope to achieve
We hope our findings will enable patients and clinicians to make informed decisions about treatments and supportive therapies for pulmonary fibrosis.