New technology to improve fetal screening
Developing new therapies and treatments for small babies and children is exceedingly difficult, which is why most drugs and medical devices are first created for adults and only later adapted for smaller patients. The Wellcome Trust has just announced that an Innovative Engineering for Health award (joint funded by the ESPRC) is being made to researchers at Kings College London to help develop new fetal imaging techniques. Meher Antia, a business analyst in the Technology Transfer Division at the Trust explains…
Babies and children are not just miniature versions of adults and there is an increasing awareness that adapting drugs and devices initially designed for adult patients for use on smaller bodies does not always produce the best results.
There are some areas of research, such as fetal medicine, that will never have analogues in the grown-up world. So rather than waiting for innovative technologies developed for adults to miniaturise, it is clearly better to ensure that they are developed specifically for fetal, neonatal and paediatric populations in the first place.
This is one reason that fetal, neonatal and paediatric care was a highlighted research area in the Innovative Engineering for Health scheme, which was jointly funded to the tune of £30 million by the Wellcome Trust and the EPSRC.
When we launched the scheme, we made it clear that we were looking for “projects with the potential to impact on particularly intractable problems in medicine or public health and exciting engineering concepts”. The team at Kings College London (KCL) got it spot on with their ambitious approach to develop life-changing technologies for a very vulnerable patient population.
Professor Reza Razavi, who led the successful bid from KCL, wants nothing less than to change the map of the UK.
The map in question is not one that shows any sort of political boundaries or geographical features, but instead shows the percentage of fetal abnormalities that are detected as part of the UK’s national screening programme.
The overall figures are already sobering – the national target for detecting fetal abnormalities is not particularly ambitious at 50%, but the sad reality is that only 35% are actually detected nationally.
What is even more startling is the regional picture (see map,). Some regions were not far off the national target (coloured in red), but others had depressingly low rates of detection – as low as 5% – 10% in the worst cases (coloured in dark blue).
The simple reason for these discrepancies is the lack of uniformly available expertise in looking at ultrasound images and making the correct diagnosis. To get a sense of the scale of the problem, there are approximately 700,000 births per year in the UK, and all these babies get an ultrasound scan at 20 weeks. The problem is that there are only 19 fetal cardiologists in the UK, making it completely impossible for all scans to get the kind of attention required to reliably detect cardiac abnormalities. If all of these scans were looked at by trained fetal cardiologists, many more cardiac abnormalities would get picked up, possibly even more than the current national target. A similar situation exists for various other abnormalities that can be picked up by ultrasound images, like lethal skeletal dysplasias or neurological disorders.
The solution to this problem, according to Professor Razavi, is to make sure that the trained fetal cardiologists concentrate on those scans that show some indication of being abnormal.
The KCL team – working with academic colleagues at Imperial College, the University of Toronto, the and the University of Florence, as well as medical device company Philips Healthcare – want to automate the screening process to achieve this.
The level of automation they want to achieve is not straightforward and nothing of this scale has been previously attempted. It starts with the need for better images.
Ultrasound images are currently acquired by a technician using a single probe, which needs to be manually moved around with one hand while the other hand operates a computer to get the best possible image. Getting ultrasound images which are good enough for a computer to pick up potentially minor differences between normal and abnormal organs simply isn’t possible with one technician trying to use a single probe.
The Kings-led collaboration are going to build a system that uses four robotically controlled probes, so that a much richer picture of the fetus can be built up and analysed.
Ultrasound technology isn’t the only area where major advances are needed. The Kings team will need to solve problems in robotics and image processing, as well build up accurate databases of fetal ultrasound and MRI images that will be used to distinguish normal organs from abnormal ones.
The entire project will take an enormous amount of time and effort, which is why the team have a budget of £10 million and seven years to bring all the pieces together.
It is easy to see how Professor Razavi’s vision for taking the lottery out of ultrasound screening would transform the lives of parents of babies with abnormalities. It would mean that parents and surgical teams could be prepared to give the baby the best possible treatment at birth instead of the hugely stressful process of a last minute scramble to treat a baby only after obvious symptoms of distress. We wish them every success.
The Innovative Engineering for Health scheme is run by the Wellcome Trust and EPSRC.