Named AI Rising Star 2020 at the recent AIMed AI Champions Awards, Dr. Addison Gearhart MD, a Fellow in Pediatric Cardiology at Boston Children’s Hospital, is quietly blazing a trail. Despite her rigorous training program, she’s heavily involved in a number of research initiatives aimed at using AI to drive innovation in the field of pediatrics


What’s been your career path – how did you get here?

I decided to become a pediatrician because as a child, I was fascinated by my grandfather’s type 1 diabetes. Over the course of my life, I was able to see how innovation directly improved his lifestyle. When he was first diagnosed his life expectancy was 30 years of age and porcine insulin was all that was available to patients with his condition. Now he is 88 years old and has a continuous glucose monitoring system. His memory is starting to go but we as a family receive alerts directly to our cellphones when his glucose levels are out of range. This technology allows him to still live independently and feel particularly loved because we are frequently calling to check in on him.


Did you ever come close to venturing into another career?

From the age of seven, I knew I wanted to become a pediatrician and I still cannot imagine a more rewarding career. My decision to become a pediatric cardiologist was a bit of a surprise to everyone in my family. I was inspired by the complexity. Each patient has unique physiology that constantly challenges me to not adopt a one-size fits all approach to treating my patients. I never go through a day without learning. It is a career that keeps you humble. My patients inspire me everyday and I am grateful for the role I get to play in their care. ​


What first drew you to tech/medicine and what do you love most about it?

Getting to see first hand how technology improved my grandfather’s quality of life and extended his lifespan provided the inspiration. I would not be here today if someone did not dare to explore new solutions to treating diabetes. There are tons of problems in medicine that need to be solved and innovative technological solutions can help bridge this gap.


What’s the best way for clinicians or people at the start of their careers to get into AI medicine/healthcare?

The best way is to network. I now live in Boston near many people and it can be intimidating to reach out to take on a project in AI when I know for a fact that I am not as well educated on the topic as the person next to me. The key is to remember that we all have valuable knowledge, insight and experience that can lead to a great project.  The first step is to recognize your importance either as a clinician who sees problems in medicine that could benefit from technology or as the data scientist who may not ‘know’ medicine but can deliver the technology to fix the problem.


Who was a great influence on your own career path?

Dr. Anjan Batra motivated me to pursue a career in pediatric cardiology. I was struck by his ability to multitask. His days were packed full managing complex patients on the floor, in clinic and in the intensive care unit, yet he made each patient feel special.  Anthony Chang motivated me to get more involved with AI. He has always encouraged me to believe in myself and to not accept failure.


Who inspires you?

My patients and their families. They fight some very difficult diagnoses with a great attitude and grace.


What drives you?

The potential to change the outcomes for my future patients.


What’s the most rewarding thing about your role?

I get to learn every day!


And the most challenging?

When my patients have a bad outcome. It’s hard to swallow that sometimes there are no available treatments. It’s hard to accept that some diseases are not curable.


How do you see the future of paediatric cardiology and AI?

This is such a hard question because there’s so much to say! Overall, I see AI as an innovative solution to address the climbing amounts of data that distract us as providers. AI can distill the data into useful trends or patterns to alert us when patients stray from the expected clinical course to act sooner or change management strategy.


Do you think the present generation of cardiologists are prepared for that future? 

We are making progress, but we still have a lot to learn. AI has embedded almost every aspect of our lives and it is no longer a question of if AI will infiltrate healthcare but, instead, when. While I am optimistic about the involvement, I think history has shown us that the involvement of the clinician will be important for the seamless integration and the development of solutions that are valuable and ethical.


What can be done better to prepare us for that future?

We need to have courses in medical school to train future clinicians. For clinicians, we could offer online fellowships, and opportunities for online teaching. I think creating a network for pairing clinicians and data scientists could also provide a practical solution.


Does the gender inequality in AI in medicine frustrate you?

It used to frustrate me, but now I see it as an opportunity to serve as a role model for future women who want to join the field.


How should we start to redress that inequality?

We need to empower more young women to pursue degrees in computer science. We need to open more opportunities for women to have roles in leadership in technology and medicine to provide role models and mentors.


What advice would you give someone starting their career in medicine?

Don’t focus on the time it takes. Often people don’t choose a particular specialty or pursue an additional degree because of how daunting it is to spend x amount of years. Life is long. You should pursue a new degree, choose the specialty you want or get more involved in a new interest. Commit to it. The happier you are in your career, the better you will do at providing care to your patients, and the more likely you are to not burn out.


What ambitions do you still have?

Where do I start? I am currently working on building a view classification model for pediatric echocardiograms. I would love to get the echocardiograms all labeled and then learn from my colleagues more about how to develop the convolutional neural network. In fact, I just started to do some coding myself.