Skip to content

UPDATE: We are delighted to announce that EPSRC have funded our CDT for five more years under the new name EPSRC Centre for Doctoral Training in Intelligent Integrated Imaging in Healthcare (i4health).

Visit our new website at:

09 Jun

Imaging and Computational Modelling in Chronic Obstructive Pulmonary Disease (COPD)

Allocated Projects

Allocated in academic year 2014-15

Chronic obstructive pulmonary disease (COPD) is a major global healthcare problem. In the UK there are over 1 million sufferers and probably 2 million undiagnosed, the majority are long term cigarette smokers but perhaps 15% are thought to occur from occupational exposure. The death rate is over 25,000 per annum often after a long and debilitating decline, and mortality is increasing in relation to other common diseases. COPD is also a major healthcare problem in the developing world from the rise in tobacco usage and the widespread use of biofuels for domestic heating and cooking. The progression of the disease remains poorly understood and varies significantly across the population, prompting interest in studying genetic predisposition and the effect on the resulting clinical phenotypes. UCL conducts world-leading research into Respiratory Disease and in particular COPD, focusing on episodes in which there is a sudden worsening of symptoms called “exacerbations”, and bronchiectasis, an associated condition where the airways of the lungs become abnormally widened, leading to a build-up of excess mucus that can make the lungs more vulnerable to infection. We are also starting work on modelling the respiratory system in studies of pollutant particle distribution in the lung. In the existing collaboration between UCL Respiratory Medicine (Hurst) and the Centre for Medical Image Computing (Hawkes, McClelland, Bragman) we have devised novel analyses of paired inspiratory-expiratory lung CT data to inform clinical phenotyping in COPD.

We are now looking to extending this work by coupling information on airway geometry down to the scale visible on CT scans (~ branching level 13 of the ~24 levels to the alveoli) coupled with image derived information on ventilation gleaned from the paired inspiratory-expiratory lung CT data. To better elucidate and understand normal respiratory function, pollutant particle distribution and disease progression we propose a coupled multi-scale imaging and physiological-modelling approach with the goal of generating patient specific models and descriptors of pulmonary function in health and disease. With these we hope to better stratify patients according to clinical phenotype by patient specific modelling of disease progression and ultimately with the goal to assign the most appropriate therapeutic regime.