Skip to content
Microsoft Industry Blogs – United Kingdom

Blogger series graphic showing an MRI scan of lungs.

This blog is a great example to show the power of cloud computing and machine learning when healthcare embraces new technologies. AI opens a world of opportunities to improve patient experience and outcomes. Connected devices and being able to collect data from various sources in the cloud means that AI can discover new insights and patterns to provide predictive analytics that change patient care.

Understanding the impact of cystic fibrosis

Around 100,000 people in the world have Cystic fibrosis (CF). Since the condition was identified around 80 years ago, improved treatments have become available, but people born with the condition today might still only expect to live to around 40 years of age.

CF affects the organs and systems, but it’s usually lung disease that shortens life expectancy. Mucus blocks up airways in the lungs, creating an ideal environment for infection and inflammation. These cycles of infection and inflammation damage the lungs so much that they ultimately stop working properly.

As a result, people living with CF need to undertake a lot of time-consuming therapy every day. On top of antibiotics, anti-inflammatory medication, mucus thinners, bronchodilators, they also need to do physiotherapy to clear their lungs.

Lung-clearing techniques and exercises can slow down the progression of CF lung disease, but these routine physiotherapy treatments are very repetitive. Children particularly can find it hard to stick to. One mother describes the difficulties to complete physiotherapy: “It was the part of the day I dreaded – blowing into a device and seeing no output. My boys just didn’t know why they had to do it, it made our lives a living hell.”

Introducing the ground-breaking work from Project Fizzyo

This is where Project Fizzyo comes in – a research collaboration between University College London (UCL), Great Ormond Street Hospital (GOSH), and Microsoft. It uses technology to gamify physiotherapy for children with CF, making it enjoyable for the children, while also collecting valuable data.

Young boy breathes into a device and plays a game on a monitor in a hospital room

Professor Eleanor Main FCSP (BSc, BA, MSc, PhD), Programme Director: UCL MSc, Diploma & Certificate in Advanced Physiotherapy talks about how the project started: “We worked with engineers and computer scientists from UCL, Microsoft and Great Ormond Street Hospital to build electronic chips for airway clearance devices and a secure and sustainable data transfer platform. We also designed computer games that are played by breathing through the chipped airway clearance device, to try and make the treatments less boring and more enjoyable.”

The results were incredible. The mother whose children tried it was amazed: “It completely changed everything. The kids became willing to do the physiotherapy as they can play games against each other. They actually look forward to it now, and that’s really taken all the stress out of the experience.”

Big data and machine learning methods to recognise patterns

By using innovative big data and machine learning methods, the project team can start to explore patterns, associations, and interactions between physiotherapy behaviours at home and clinical outcomes for young people with CF. The project can also find out whether gaming during treatments will make treatments easier to do more regularly, and whether this helps children and young people to stay well.

A device to capture data is clipped onto a persona

Professor Main goes on to explain: “With our software platform, we capture large amounts of breathing data from children and young people doing their routine physiotherapy sessions every day. We also get data from their Fitbit activity trackers. The data for each child is rich and unique, giving us the opportunity to apply AI machine learning solutions to different aspects of Project Fizzyo.

“For example, physiotherapy treatments are done using different devices, usually selected on the basis of age or personal preference amongst children with CF. We applied machine learning to identify what kind of device was being used for the physiotherapy treatment by analysing the data for individual breaths. We also want to use ‘cluster analysis’ based on children’s daily physical activity and airway clearance profiles to see how different ways of doing physiotherapy treatments at home impact health.”

Looking to the future

In the long-term, Project Fizzyo hopes to be able to know more about which airway clearance treatments are the most effective, and how often they need to be done. This may also help clinicians get better at predicting when infections are likely to happen so children can receive effective treatments earlier.

“With the data pouring in every day, we’re already staggered by the fascinating insights we’ve had into the way children engage with their physiotherapy treatments and physical activity at home.”

– Lee Stott, Microsoft Evangelist

This kind of work has the potential to radically change the way care is delivered and can evaluate the effects of medical interventions like physiotherapy. It could facilitate personalised and evolving care plans for children as they grow, and these plans could change with life circumstances.

Project Fizzyo video

This really is a fantastic story of how the intelligent cloud and connected devices can radically transform patient experience and tolerance of treatments for children with CF. I can’t wait to see how it develops – and the potential it has to positively impact the care, treatment, and research in these genetic diseases.

Learn more

Kelly Limonte headshotAbout the author

Kelly is the Healthcare Industry Manager at Microsoft UK, working with transformational digital partners and NHS customers to pilot solutions for collaborative working and empowering everyone to do more. She has 15 years’ experience working alongside the NHS, and is passionate about the power technology has to create positive change in healthcare.