###Characterising early lung fibrosis using advanced image analysis
Project ID: 2228bd1071 (You will need this ID for your application)
Research Theme: Healthcare Technologies
UCL Lead department: Division of Medicine
Lead Supervisor: Sam Janes
Project Summary:
Respiratory disease affects 1 in 5 people in England, and even before COVID-19, was the UKs third biggest cause of death costing society £9.9 billion annually. Lung fibrosis affects 1% of patients over the age of 60, accounted for 0.9% of all UK deaths and is increasing in incidence and prevalence as the population ages.
Patients with lung fibrosis are typically diagnosed at a late stage. Early diagnosis can only be performed with imaging as lung breathing tests will show normal values in early disease. On clinically-acquired lung imaging (computed tomography scans) however we still do not know what early fibrosis looks like as we do not have histopathological scale ground truth of early disease.
A new imaging modality called hierarchical phase contrast tomography (HiP-CT) is able to capture 3D histopathological scale imaging of the lungs of patients with early fibrosis through a collaboration between UCL, the European Synchrotron Research Facility and Antwerp University. Supported by a $5 million award by the Chan Zuckerberg Initiative, lung imaging data will come to UCL for analysis.
Our project aims to employ novel semi-supervised deep learning techniques to allow labelling of this extremely high resolution imaging data. This will allow us to understand for the first time the pathophysiological changes that manifest at the earliest stages of lung fibrosis.
We will also map ground truth features identified on HiP-CT to clinical CT images in patients at high risk of lung fibrosis using generative diffusion models. This will develop the first imaging biomarkers of early lung fibrosis.
Currently the mean diagnostic delay in a patient with lung fibrosis is 2.1 years, with mean survival between 3-5 years. Bringing about early diagnosis of lung fibrosis using HiP-CT insights could help develop drugs that slow disease progression transforming the outlook for lung fibrosis patients.