June 1, 2013

Saving Lives of Premature Babies with Real-Time Analytics

Isaac Lopez

Big data techniques may hold the key to help the littlest of people, says Carolyn McGregor, the Canada Research Chair in Health Informatics at the Ontario Institute of Technology, who says that real-time data analytics may hold keys to discovery in neonatal healthcare.

McGregor got her start programming executive information systems for a bank in Australia, which she says was among the earliest analytics systems at the time. She expanded her expertise to retail and doing consulting work for a number of years until a meeting with a neonatologist put her on a new course.

“He had come to the university with a wish,” she explains. “That he be able to use every breath and every heartbeat, rather than just having information summarized down to a reading every hour. McGregor says that the problem from a computing context was very similar, whether trying to understand customer or supplier behavior to trying to understand the behavior of premature babies.

The issue struck home for McGregor when a chromosomal abnormality impacted the loss of her own prematurely born daughter. “I came to realize that in the case of my own daughter that nothing could have happened to change the outcome for her, because her destiny was right in that very first cell divide, but there are so many other babies in the neonatal intensive care unit setting who have the potential for a wonderful outcome in their own life and with their own family.”

McGregor says the event motivated her to take the programming skills and techniques that she developed in banking, retail, and finance, and apply them in a domain where she could save the lives of other prematurely born babies.

“Every breath and every heartbeat in those prematurely babies matters, and we need to watch them, and we need to support them as they grow,” says McGregor explaining the link between big data and prematurely born babies. The babies are meant to be in amniotic fluid, she explains, which means that their development is happening in a stress inducing environment that they weren’t intended to.

With all the monitors hooked up to these premature infants, advanced computing is able to provide analytics which give early indicators of problems that may be developing before it’s too late.

“What we’re starting to discover is that if we use advanced computing tools to watch every beat of the heart and every breath, and analyze that, we can start to see when a baby is becoming unwell because they’re starting to have to fight off an infection, we can start to see the impact of oxygen levels that could potentially impact the development of the eye, we can start to see the impact of pain management drugs on the body, and a range of different things.”

Addressing the data, McGregor says that a baby’s heart beats more than 7,000 times in an hour; they breathe more than 2000 times; their blood oxygen levels are taken 3,600 times in an hour. Prior to the development of these methods, only one number was written down each hour for each of these vital statistics.

McGregor says that the analytics work that they’ve done on the premature babies has great applicability to other healthcare spaces, including such case studies as infection, chronic condition patients, ambulatory care, cancer, leukemia, diabetes, or any chronic type condition – in many cases from in-home care situations.

“We have a platform that allows you to send that medical data through the internet and have it be processed in the cloud, which has enormous potential in the future for clinical care.”

Related Items:

Biotech Gets a Boost With New UK Big Data Institute 

Detecting Cancerous RNA Signatures with Big Data 

Preventing Brain Injuries with Predictive Analytics 

Share This