Enhancing the characterisation of early respiratory disease to improve population health

I aim to transform our understanding of the anatomic processes underlying the earliest stages of chronic obstructive pulmonary disease (COPD) and lung fibrosis, two globally important diseases.

Employing Hierarchical Phase Contrast Tomography (HiP-CT) to image entire lungs at 25μm resolution from two European centres with accessible correlative histopathology, I will develop whole-lung computational segmentations of vessels and airways characterising lung morphology across health, early- and late-stage COPD and fibrosis.

Histopathology-informed HiP-CT airway and vessel maps will allow accurate labelling of contemporaneous whole-lung micro-CT imaging.

Advanced artificial intelligence techniques (generative diffusion models) will map micro-CT and clinical CT allowing super-resolution (micro-CT scale) insights on clinical CT imaging.

Interpreting super-resolution clinical CTs will inform a new vocabulary of radiological terms describing patterns of damage in early and progressive COPD and fibrosis.

This will democratise our technological insights to radiologists and clinicians around the world without access to the same technology.

Computationally-derived imaging biomarkers informed by tissue microstructure will also be applied to longitudinal clinical CT imaging in four UK lung cancer screening populations (over 30,000 subjects).

The biomarkers should result in quantifiable and interpretable prognostic and theranostic biomarkers specific to early COPD and fibrosis and allow trial cohort enrichment.