King's College London

A big part of leading a multi-disciplinary research team is solving problems, which is also a key part of startups and entrepreneurship.

For example, the first team I was part of at King’s moved to another university. I had the option to go with them but decided to stay. Suddenly there was nothing much left behind, but having had the opportunity to see what worked meant we could look to rebuild a multi-disciplinary team as well as bring in people who were dedicated to the translational and commercial aspects of research and how that translated into improved patient care.

We had clinical, engineering, computer sciences, and even chemistry staff, all working together. The next ingredient was working with industry and building strong relationships with companies so research would translate into their healthcare products. More recently also extending this to entrepreneurship and supporting our researchers in turning their ideas into start-up companies.

This was the recipe we used to start growing the School (then the Division) of Biomedical Engineering & Imaging Sciences and building a more porous research culture at King’s. I then moved on to research leadership roles across all of Kings and Kings Health Partners to spread our recipe for success more widely.

Essentially, my move to Fraiya came about from the leadership roles where I was supporting others to do all of this, and over time found myself wanting also to have a go at it too!

Can you tell us more about Fraiya, its technology and also its mission?

Following a grant call from the Welcome Trust and EPSRC we started by taking a long look at problems that were difficult to solve, and where new technological approaches might make a real difference to patient outcomes.

I’m a children’s cardiologist, treating heart problems in babies, so it felt natural to start in this area. When we began looking through the lens of technology and questioning established practices and processes, we soon identified an issue sitting in plain sight.

Put simply, we have a much better chance of successfully treating a baby born with a congenital heart defect if clinicians are aware of potential issues before a baby is born. This way we can look for babies to be born in a hospital with specialist experts who can treat them from birth, including major operations to try and help improve or fix their hearts, rather than having very sick babies presenting to a local hospital without a diagnosis and all the time deteriorating while being transferred to a specialist centre for diagnosis and treatment.

We know from studies that babies who are diagnosed by the routine pregnancy ultrasound scans have better outcomes, are less sick as babies, with for example shorter stay in intensive care and longer term have less complications too. Heart issues in a newborn can affect other organs in their bodies as well, including the brain, so even where we were able to treat patients successfully you might still see issues with learning or behavioural difficulties in a child over the years.

Picking up congenital problems during pregnancy is one of the important reasons why we do ultrasound scans at around 18 to 20 weeks as a routine for every baby. However, as these congenital problems are made up of over 100 different conditions that are individually rare (all together they make up 2% of pregnancies) these scans can sometimes miss the problem. For example, with congenital heart problems, often the most serious and making up half of all abnormalities, only around fifty per cent are picked up.

The traditional approach was to see better training for ultrasound sonographers as the solution, but we saw this instead as a challenge where advanced imaging technologies might help solve the problem. These scans are hard, and we can’t ask a fetus to hold still while we get a better look at them during the scan! A person doing a scan might not have encountered one of these rare conditions in a year or two, or if they are relatively early in their career never at all. So even if their training is at the highest possible level, the chance of missing an issue is still there and improving training can only take us so far.

The Fraiya technology counters this by adding an AI element to the scan. The AI firstly sorts out all the images into those that show one of the many views that a sonographer needs to look at during the scan as well as automatically measuring the parts of the foetus that have to be monitored and provides this information back to the sonographer in real-time. It also turns the scan from an unstructured scramble of images into a structured study that can have a second review by clinical experts assisted by AI disease detection algorithms.

In other words, we’re not replacing the expertise of clinicians but rather enhancing it with new tools, allowing them to focus on care when they are with a patient, and helping improve their diagnostic accuracy by allowing a separate review of the data.

It also means we can cut the average scan time from around 20 minutes to something like 11 minutes, maximising a sonographer’s productivity and increasing the number of obstetric scans without compromising on care.

Finally, Fraiya is now a resident at the London Institute for Healthcare Engineering and also a part of its Venture Builder programme. What benefits have you and the team gained from joining the LIHE MedTech ecosystem? 

LIHE is a fantastic environment for a new start-up with not only great space but with the Venture Builder great support in the many aspects of getting your company of the ground that would be hard to get on your own. This Includes quality management and regulatory support, clinical trials and health economic expertise, business planning and marketing help and assistance with external communications. All so valuable!

Also, I think when you go into commercialisation and startups, you have to realise that it’s not always going to work out. Of course, you want something to be a success, but the learnings we gain when something doesn’t work can be just as valuable. One of the most attractive things about LIHE is being based among other companies on the same sort of journey, so you can learn from each other and not have to make every mistake for yourself!

The clinical engagement that comes from being based on a hospital campus is also so valuable. At the end of the day, you want to be the company that gets its idea to market. This means not only having a great technological solution, but also something that works for its users, is robust, and that addresses the real problems clinicians are facing rather than being a technological fix in search of a problem to solve.

Everyone in the LIHE community is so generous about sharing their learning, their experience, contacts and connections, and helping each other, so we have the opportunity to contribute to and build a MedTech ecosystem around both LIHE’s provided experts and also peer support and shared learning.