In 2020 the Charity awarded Dr Tom Burgoyne with a research fellowship that would support him for up to 12 months of research into primary ciliary dyskinesia (PCD), with a focus on how new technology will help diagnosis.

What is PCD?

PCD is a rare genetic respiratory disorder that affects the patient’s cilia, which are small hair like structures in the airways. Mucus is produced in the airways which traps dirt and germs. Normally these tiny cilia hairs move to brush these away from the airway, but abnormal cilia may not move as they should. This creates a build-up of mucus and potentially harmful debris, that might lead to infections, breathing difficulties and other symptoms.

PCD is difficult to diagnose, but early diagnosis and treatment can stop PCD causing permanent lung damage.

My work is about having a better understanding of this very complicated disease so we can get to a point where we can diagnose faster and provide better care for patients

Tom’s research

In his work, Tom uses a powerful electron microscope to examine patient samples. Images 

are taken using the microscope and until recently these images were analysed by eye to pick up any abnormalities. To reduce the time it takes to diagnose PCD and reduce costs, Tom has been finding ways to automate collection and analysis of these images with machine learning.

The team have been working on automating the analysis of images since 2019. “We've been putting the images taken with the electron microscope into the computer system and training it to recognize when the cilia look normal and when they look abnormal,” said Tom. “One of the problems is that there was no way yet to automate collection of the images.”

The aim of Tom’s work during this fellowship was to setup the electron microscope to recognise cilia and then automatically collect images that can then be used for the machine learning system. While the microscope can recognise the cells, recognition of the cilia, which are even smaller structures, is proving to be difficult, but Tom is making strides towards this goal.

What’s next?

This innovative research has helped Tom secure further funding from the National Institute for Health and Care Research. Tom’s research, coupled with advancements in tech and machine learning, is leading to a deeper understanding of PCD, as well as providing insights that can be used to better understand and diagnose other lung conditions.

The machine learning software has been successfully trained from electron microscopy images of cilia and from trials it can reliably assist in diagnosing PCD. “We compare results with the software to those produced on the normal diagnostic pathway. At some point we want to integrate the system fully into diagnostics service and then it will really help to benefit patients even more.”

“I have quite a lot of projects on the horizon,” Tom said. “With regards to the machine learning, we want to roll that out to other centres in the UK and potentially beyond that, globally. We are also interested in looking beyond PCD and examining video data from patients with, for example, cystic fibrosis and bronchiectasis, or infections with the aim to teach these machines to pick up both respiratory disorder and infections too.”

Software written by Tom for image analysis has been made freely available for use, especially at diagnostic centres that specialise in PCD. Some programmes have been downloaded over 900 times. “This software is now used routinely alongside the normal diagnostic setup, especially for patients who have more difficult to detect defects.

Tom will continue his research looking at PCD and its diagnosis. We wish him all the best in this research that could help patients with PCD achieve a better quality of life.

It is through the generosity of our donors and supporters that we can fund essential research to not only improve outcomes for patients in our hospitals, but also inform the treatment of patients with lung disease everywhere. Play your part in our mission to treat and beat lung disease.