Artificial intelligence and robotics are becoming important factors in the progress of healthcare, thanks to pioneers like Dr. Carolyn McGregor AM, the associate dean, research, at Ontario Tech University. (She’s also the research chair in Artificial Intelligence for Health and Wellness, the director of the Joint Research Centre in AI for Health and Wellness, and a two-time Canada Research Chair in Health Informatics. No wonder we were so excited to chat with her!) But Dr. McGregor didn’t set out to change the world of AI in health and wellness when she first joined the workforce after her post-secondary education; she started her career, 30 years ago, as a data scientist in the banking industry in her home country of Australia. So how did she end up becoming one of the foremost minds on health informatics and living on the other side of the world? Read on to find out.

I’ve interviewed many noteworthy people over my 15-plus years as a journalist — TV celebrities, politicians, famous authors, even Meghan Markle before she was the Duchess of Sussex (she was so nice!) — and Dr. Carolyn McGregor ranks up there with one of the most interesting people I’ve ever had the privilege to speak with. That’s probably why I had so many false starts when I started to draft this article; I felt like I was failing to convey Dr. McGregor’s passion for her work, and her fascinating personal and professional history. So, in the end, I’ve decided to let Dr. McGregor tell you about her life and career herself, in her own words.

Katie Dupuis for The Spark: Tell me a little bit about your origin story and how you came to work in this field.

Dr. Carolyn McGregor: I started out in the analytics area when it was just forming. I was living in Australia and I was working for a bank at the time. In that job, I was asked to work on the first analytics platform the bank had. I was then headhunted to do something similar for one of the largest retail chains in Australia. Eventually, I ended up working as a consultant to help a number of organizations on their strategies, on how to start to understand their businesses through analytics. But I was also working with a university part-time, and a neonatologist came to the school through a different department and said, “Look, we’ve got all these numbers that show up on medical devices, and I really think that there are some subtle messages in those second-to- second stats. I’d really like to work with someone on this.” I could see that, while I was looking at patterns in how people spend and save their money, or what they were buying in the grocery store, working to understand premature infants was much more complex. But, it was also a similar style of problem. We didn’t have a computing system that allowed you to constantly receive heart rate and breathing and all of this information from the body. Having that could mean watching for changes, and if something was changing — like there were signs that the baby was developing an infection, or could be at risk for a hemorrhage — physicians could determine a course of action.

The Spark: What a change from the initial application of your skill set! Did you find it rewarding?

McGregor: Well, I had a personal connection. When I met the neonatologist, I was pregnant with my first daughter. Unfortunately, she was born premature. She had a rare chromosome abnormality and she passed away. It just made me realize that I wanted to focus my energy on helping families bring their babies home. That’s really what started my journey. Then I met the founding dean for Business and IT at Ontario Tech at a conference in Hong Kong, and she said, “Would you consider moving to another country?” I said yes, and I was awarded the Canada Research Chair to come over.

The Spark: What did you do here when you first arrived?

McGregor: I worked for many years with the Hospital for Sick Children — we deployed an initial system there. I’m now working with McMaster Children’s Hospital. But other people could also see opportunities outside of neonates, and I was introduced to retired astronaut Dave Williams. He could see the potential of my work to monitor the physiology of astronauts. I was then invited to the Canadian Space Agency. They introduced me to NASA and to the Russian Institute of Biomedical Problems, and through that connection, I had the amazing opportunity to work with founding space cardiologist Roman Baevsky. He was Yuri Gagarin’s cardiologist for that first space flight.

The Spark: So, babies and astronauts?!

McGregor: I get that a lot. But a baby floats around in the womb, and their bodies change when they’re born. In space, you’re dealing with microgravity and how the body changes there. It’s the reverse.

The Spark: What other applications of your work have come to the forefront?

McGregor: Well, at one point I met the staff sergeant of a tactical police team, and he said that we should be monitoring tactical officers. Their behaviour is observed but we don’t really understand their physiology [in stressful situations]. So I was able to learn what they knew about physiology for military and tactical personnel. We know if their heart rate gets too high, for example, it impacts their fine motor skills, their cognitive skills, etc. With that in mind, my research has essentially become a way to monitor and assist people who are in extreme conditions.

The Spark: Wow. That’s incredible. But let’s take a step back. If your background is in analytics, how did you learn so much about the human body in order to apply the tools you’ve developed?

McGregor: There was a lot of on-the-job learning. It’s funny, people will hear me speak about astronaut or neonate physiology and they’ll ask, “Were you a doctor first? Or a computer scientist first?” But I haven’t actually completed a degree in anatomy or physiology or medicine. I’ve just spent time with people and learned what they needed.

The Spark: Can the work you do apply to mental health as well?

McGregor: Absolutely. Interestingly, it has been shown that heart rate and breathing behaviours change in association with mental health. Depression, anxiety, PTSD…they all have an impact on heart rate, breathing, cardiovascular. So we are working in the wellness space for professions like astronauts, tactical officers and firefighters, but I’m also now working with Ontario Shores to look at ways to help staff. They often have to deal with aggressive behaviours in patients, and it can cause post- traumatic stress injury. We’re performing a study at the moment where we have an actor who behaves like an aggressive dementia patient, and we measure the physiology of staff when they are trying to de-escalate the scenario.

The Spark: So, really, your work is wide reaching. You’re measuring physiological changes in people on the job but you’re also able to use the physiological signs for training.

McGregor: Yes. And if we can repeat-expose people to the training in a safe space — everyone has different triggers depending on their life history — we can individualize the skill development.

The Spark:That’s really cool, but I imagine there are so many factors to consider.

McGregor: That’s true of all my work. We have to consider bias in all populations — neonate, astronaut, etc. We have to consider age, race, background. We also have to consider the fact that AI can cause harm, just like any other medical intervention can cause harm. We also have to think about people’s perception of artificial intelligence. You really have to understand what it isn’t doing, as much as what it is doing. Knowing what it isn’t doing helps you to understand the scope of your own responsibility.

The Spark: In terms of the technology, how would you chart the trajectory of the first systems you worked on 30 years ago to now?

McGregor: In the early years of my career, in Australia, there were no systems that had the complexity of watching different signals, logging different features. The closest thing we found was cargo ships. When they are loading and unloading equipment, they have to offload liquid from the ballasts to keep the ship steady. That was the closest computing system at the time that we could build off of. Then I met some researchers at IBM who were using stream computing. Using that theory allowed us to create a way in real time to gather the very first piece of information. Then I patented a new data mining approach that I was able to apply to a number of neonatal conditions. And it’s not a black box, it’s a white box — we know what it’s finding. But it did take us a few years to demonstrate that these systems were reliable and stable in different domains.

The Spark: So how do you find new opportunities to try your work?

McGregor: Sometimes people come to me, and sometimes I go to them. I’m comfortable with failing. I don’t see it as failure. I see it as, “Well, trying that way didn’t work.” I think that’s a really important message for people to understand. When you’re innovating and inventing, it’s really about trying, learning, revisiting. I use the term “fail” because I want to normalize the language — you just have to keep trying.

Robot Surgery

Artificial intelligence is finding its place in many different ways in Durham Region — including our local hospitals. Last year, Lakeridge Health Oshawa introduced the da Vinci surgical robot, a $5.1 million-dollar piece of equipment that is known to “lower pain scales for patients, result in shorter lengths of stay, lower complication rates and lower rates of readmission,” said Dr. John Dickie, the chief of surgery at Lakeridge Health. While the technology isn’t new, there weren’t any robotic surgical options between Toronto and Kingston prior to the purchase of the Oshawa unit. The portable robot has four surgical arms with integrated cameras, which allows the surgeon to see the field of view in more detail. The da Vinci’s hands also have more flexibility and range of motion than a human hand.

Recommended Posts