Facebook Pixel Code
Banner image
About the Institute

Profile of Melissa McCradden

Photo of Melissa McCradden
Dr. Melissa McCradden

By: Maia Leggott

Melissa McCradden, PhD

  • Post-doctoral Fellow in Ethics of AI in Healthcare, Department of Bioethics and Genetics & Genome Biology at The Hospital for Sick Children and Vector Institute for Artificial Intelligence

Where are you from and where did you study?
I was born in Toronto and spent most of my childhood in London, Ontario before returning to the city 13 years ago. I did undergraduate studies in psychology at York University because I was interested in understanding human behaviour on a neuronal level. I then completed my PhD in neuroscience at McMaster University focusing on the neuropsychiatry of brain injuries.

I’m currently pursuing a Masters of Bioethics with the University of Toronto, which represents a fascinating intersection of philosophy, medicine, and people’s lived experiences. A neuroscience background lends itself well to this research area; there are a lot of similarities so I’m grateful for that background as I explore this new terrain.

What is it you're researching right now?
Oh, there’s a lot [laughs]. It’s an incredible research area, but mainly my interest is in the clinical space, and how AI-assisted technologies interact with patients and care providers. It’s exciting because a role like this hasn’t existed before. 

My research combines the work of Randi Zlotnik Shaul, Director of Bioethics at SickKids, with that of Anna Goldenberg, Senior Scientist in Genetics and Genome Biology. It focuses on “explainable AI,” which includes AI techniques that can be easily understood and trusted by humans. This work supports ethical decision making with AI-assisted systems by identifying appropriate accountability measures and integrating core paediatric bioethics concepts, such as shared decision making and child- and family-centred care. 

From a computer science perspective, the focus is on how an algorithm works and what features are present. On the other hand, the clinical focus is on patient care and how AI technology can complement that. Clinicians want to understand what`s relevant for them to explain a decision or recommendation to a patient and their family. Our working hypothesis is that the greatest patient benefit comes from using explainable AI in ways that are both clinically and statistically relevant.

Why do you think this is an important research area to focus on now?
For some time, the computer science community has noted the ethical implications of the work that they’re doing, so that community has led the charge in requesting ethics input. Right now, we’re on the verge of understanding how effective and helpful these AI methodologies can be, but we are still figuring out how to implement them in a way that’s clinically meaningful and ethically sound. 

Another consideration is the adoption of electronic health records systems and how they enable the efficiencies of different AI techniques. There’s been a recognition in the computer science community that their solutions don’t necessarily tap into the broader ethics literature that exists. This ethics knowledge base drives concepts that we now recognize as central to modern medicine, like patient-centred care and respecting patient/child autonomy. 

These concepts may not be familiar to the computer science community. Similarly, ethics experts and researchers may not be familiar with the computational mechanisms used in this research. The vision Randi and Anna had for my fellowship was to develop a specific expertise that has a solid technical grounding in the methodologies that are used, but that also builds capacity, awareness and in-depth knowledge of the broader bioethics literature.

Why do this research at SickKids?
I’ve wanted to be at SickKids my entire life; it’s a very special place. SickKids is leading in this research area, and one of the greatest things about working here is the incredible atmosphere of collaboration. Clinicians are enthusiastic about our research, computer science students are engaged and supportive, and I think these, coupled with other factors, mean there is a great opportunity to lead the charge in this field. 

On a personal level, my spouse has a severe peanut allergy; when he was young, he had a serious allergic reaction and was airlifted here, where he was in a coma for several days. If it weren’t for SickKids, he might not be here right now. I owe my family’s existence to SickKids.

Would you say that inspires your work?
Absolutely. Also, this field is the ideal intersection of a lot of my interests, like philosophy and ethics, research methodologies and technical understanding, as well as what constitutes good clinical care. In the past these interests were all siloed, but this fellowship integrates all of these ideas.

Did you always love science?
The thing that drew me to science was psychology and understanding behaviour motivations. I have a very specific interest in psychiatry, which like AI, integrates a lot of different ideas, including theories about the self and personal understanding of the world. I chose neuroscience to try and understand how behaviour happens at a neuronal level and how to find solutions for behavioural issues that cause personal distress. Under a clinical care lens, there’s a lot of complex care and considerations in psychiatry.

Who is your favourite scientist?
Can I say my supervisors? [laughs] But in all seriousness, my PhD supervisors, Dr. Patricia Rosebush and Dr. Michael Mazurek, had an incredible ability to integrate their understanding of a patient`s clinical experiences with research methodology and basic science knowledge in a way that really furthered our understanding of best practices for patient care. They taught me how to really think critically about everything that I did in a research context, and to think carefully about what the effects of certain knowledge can be.

What would you say is the most controversial question in your field?
AI is now in everything we do; it’s become an everyday term that holds a lot of influence over the decisions we make. In the public sphere, the term has become associated with, say, the Terminator [laughs], versus the clinical decision support systems that we work with in AI in medicine. I think this presents a real challenge to patients’ democratic participation in decision-making. 

The weight of the term “artificial intelligence” often colours how you perceive certain information. It can be very challenging to decide what you think about the use of AI, how you feel about it, and how you’re going to proceed in light of information that comes from this system that can seem scary at first.

What are your major interests outside the lab?
My family. I have two kids; my daughter is almost two and my son is almost four. My spouse and I were both athletes, so we are a huge sports family—we were devastated when the Leafs lost in the playoffs again this year! My athletic background is in diving and I’m still very active in that community. Most of our free time as a family is spent on different activities and athletics – my son recently learned how to ride a bike which is very exciting!

Very! Speaking of exciting, are you reading anything of note lately?
Something I’m really interested in right now is taking a feminist perspective on machine learning. There’s a great online publication about feminism in data science that I’ve really been enjoying. It’s encouraged me to look back to some of the feminist philosophy of scientific approaches and consider how some of the concepts that have been presented with a feminist lens can apply to machine learning.

If you could give one piece of advice to someone considering a research career, what would it be?
I think all research is very exciting, and I like the direction it’s been taking recently where the focus is more on research that is meaningful to patients, to people and to the community in general. If I was going to advise someone about a career in research, I would say pursue something that has personal meaning to you, because if that’s the case then it will be meaningful to others as well.

What does the Peter Gilgan Centre for Research and Learning mean to you?
Being here gives me the opportunity to work more closely with the computer scientists who do this work, and to learn from them. A lot of what I do crosses the worlds of ethics and computer science; the two areas are highly integrated, so a lot of my time is spent in that intersection. 

The PGCRL was designed in a way to facilitate collaboration; it offers a centralized place where I can work closely with both computer scientists and bioethicists. Whenever I talk about my research people are very excited about the vision for this fellowship, and there’s a strong feeling that this kind of knowledge is going to be very important moving forward. 

Melissa is currently working on creating an opportunity at SickKids to generate discussion and exploration of the ethical issues surrounding clinical AI projects at the hospital. Please contact Melissa if you're interested in learning more!

May 2019