Research in big data analytics working to save lives of premature babies
TORONTO – It’s hard to imagine that something so small could create enough data to be used in research for big data analytics, but researchers at Toronto’s Hospital for Sick Children are using premature babies to do just that.
‘Big data analytics’ is the study of a large amount of data sets, broken down by computer systems that can detect trends. Researchers are collecting high-frequency physiological data, including heart rate and respiration rate, to create algorithms that can predict when a baby is at risk of infection and other health complications.
Premature babies, or preemies, demonstrate tell-tale signs of infection in changes in their heart rate up to 24 hours before an infection takes hold.
Once put in place, these algorithms can help doctors walk the thin line between life and death that preemies face.
“About a quarter of babies who are born premature will end up with infections and ten per cent of those will die – that’s quite a significant problem,” said lead researcher Dr. Carolyn McGregor, Canada Research Chair in Health Informatics at the University of Ontario Institute of Technology.
“When you think about the fact that there is 7,000 heart beats per hour, and more than 2,000 breaths in the hour, then you can start to see that there is a lot of data to a baby.”
McGregor, who began her research in this field back in her homeland of Australia in 1999, wanted to find a way to break down the flood of information coming from these infants so that doctors were able to diagnose conditions faster and more effectively.
“What they need is a proper computing platform that allows them to make sense of all of that high-frequency data in a way that can help them help the babies even more,” McGregor told Global News.
After being awarded the Canada Research Chair in Health Informatics position in 2007, McGregor got to work on the cutting-edge program, called ‘Artemis,’ in Toronto.
The program currently has about 1,000 babies enrolled from the Hospital for Sick Children, alongside a cloud computing version of Artemis that runs out of Women and Infants Hospital in Providence, Rhode Island.
The cloud-computing-based version of the program, which is also in place at a children’s hospital in Shanghai, allows the hospital’s machines to be hooked up to Artemis’ network by cables.
McGregor and her team then process the information remotely at her lab in Oshawa, Ont.
But McGregor believes that using cloud computing in the research phase of the project will shed light on the usefulness of this type of technology for preemies in rural areas.
“What often happens is that a baby may be born at term, or slightly premature, and then they start to become very unwell or problematic. Then they would get in touch with a neonatologist and potentially arrange transportation for the baby,” explained McGregor.
“Whereas with our system we could be letting the local pediatricians know that we are picking up something that is starting to happen with that baby and that they can seek support earlier – or, potentially even stop the need to have that transport because they have intervened earlier.”
The technology could prove very useful for expectant mothers in rural Canada.
A report released Thursday revealed that some women had to drive more than two hours to reach a hospital to deliver their babies.
The report, done by the Canadian Institutes for Health Information, also found that women from rural areas were more likely than urban women to have their babies delivered by a family physician and less likely to have the birth attended by an obstetrician, gynecologist, or midwife.
This could mean that rural women and their babies may not have access to a neonatal unit where they delivered – which is where McGregor’s technology may come in handy.
“They don’t have the tools to say – well the heart rate fell to this level, it lasted this long, and the blood oxygen also fell – they aren’t able to richly describe it,” said McGregor.
“We’ve created an environment to automate that so that they can get much better information about that clinically, which will help them to start to work out what’s going on.”
McGregor and her team partnered with IBM Canada’s research and development centre, and is working to build a robust computing platform that can eventually be made available to more hospitals.
She hopes that Artemis will add to the ever-changing landscape of healthcare technology and, more importantly, save some young lives.
“There has been a lot of research and talk about how genomics has quite the potential in healthcare. I think doing this high-frequency analysis on peoples bodies has as much, if not more, potential to really change the way that health care is provided and supported,” said McGregor.
According to numbers from the March of Dimes released in 2009, roughly eight per cent of babies born in Canada are premature.
© 2013 Shaw Media