This startup detects pain in premature babies and alerts clinicians in real time
About 1 in 13 babies in the UK is born prematurely — that’s almost 60,000 per year — and data shows the number is on the rise, according to UK-based charity Bliss.
Premature babies are unable to verbally communicate how they’re feeling and they undergo around 10–12 potentially painful procedures a day for an average of 50 days. Pain assessments are still carried out manually, they are time consuming and mostly paper-based, making it difficult for nurses and clinicians to ensure consistent monitoring and accuracy.
For Arthur Howard, a PhD student in Behaviour Informatics at Newcastle University with a passion for machine learning, “effective pain management is crucial”, as exposure to high levels of pain or medication at such a critical period of neurological development can have serious developmental outcomes in later life.
Part of the Conception X Northern Deeptech Hub, Howard’s startup Affectronics is building a deep learning system to provide 24/7 pain monitoring and alert clinicians to the earliest signs of discomfort in premature babies — in real time.
The technology combines audiovisual data — such as face expressions, body movements and vocalisations recorded by cameras and microphones — with the physiological signals picked up by vital signs monitors, and it classifies pain through a score from the Neonatal Pain, Agitation and Sedation Scale.
“There are currently no solutions in the market that use all the data available to us in an affordable and non-invasive manner,” Howard says.
“It’s a rapidly changing field. Not too long ago, just before the 80s, they thought premature babies’ nervous systems were so underdeveloped that they didn’t experience pain at all, and it would be common practice to carry out routine surgeries without anaesthesia. Then, the scientific consensus shifted, and the belief is that they definitely do experience pain, maybe more acutely than adults because they can’t localise it. But it’s very early stage stuff, and something that’s still being explored.”
For the past few months, Howard has been running a clinical trial at the Royal Victoria Infirmary in Newcastle, collecting and building a well annotated dataset to feed his deep learning algorithms, and he’s recently been invited to present his initial findings at the International Symposium on Pediatric Pain in Canada later this year.
Asked what success would look like, Howard says it would be “a system that works well enough to be clinically useful”.
“One of the main challenges is that you can’t open source patient data, so we will have to collect a big enough dataset to robustly train the models, carrying out similar trials elsewhere to make sure demographic distribution is well balanced,” he adds.
Howard first joined Conception X to meet other PhD students working in the same space and with similar aspirations based in London and the South, and came to realise he wanted to build a startup while on the programme — “to innovate and actually make stuff that people care about, not just research”.
“My network absolutely exploded off the back of Conception X,” he says. “It was nice to feel that sense of there being an innovation ecosystem that does connect across the whole country, and it’s not just siloed everywhere. Meeting so many exceptionally intelligent people was such a unique experience that will stay with me forever.”
One of the most valuable lessons for Howard has been the adoption of the lean methodology, which has helped him experiment, move faster and ultimately get more done — in his PhD work as well.
“Even if I wasn’t necessarily looking to move Affectronics forward as a business, those are ways of working that academics can really benefit from,” he says. “It’s made me very process-focused, and the results are showing.”
Arthur Howard joined Conception X as part of the inaugural Northern Deeptech Hub cohort, launched in partnership with Newcastle University, Durham University and Northumbria University to connect aspiring PhD founders in the North East with the local deeptech network, help them gain business skills, access funding and launch deeptech startups. As of 2023, the Hub includes Teesside University.