Emilie Bruzelius has done something uncommonly interesting with her professional life: created a career that mixes epidemiology and data science.
Bruzelius works as a data scientist and epidemiologist at the Mount Sinai Institute of Epidemiological Health, where she uses a brew of statistics, data science and machine learning to combat inequities in global healthcare. She’s pursuing her Ph.D. in epidemiology at Columbia while working on the data-science team at Arnhold Institute for Global Health at Mount Sinai. She’s also a graduate of DSI’s certification program
Originally from Boston, Bruzelius studied biology and sociology as an undergraduate at Brandeis University, after which she intended to apply to medical school. Instead, she attended Columbia University’s Mailman School of Public Health, where she earned a master’s degree in public health. In this Q&A, she answers a few questions about her pioneering work in public health.
***********
How did you get interested in studying at the Data Science Institute?
After my master’s in public health, I worked at Columbia. I was also working with big data analysis in electronic medical records at Mount Sinai and that’s where I got to know about the Data Science Institute.
Did you get a lot out of the DSI certification program?
I had a mix of great experiences. I loved my data-visualization class. That was where I learnt how to code; I learnt everything all at once. One of the things that I appreciate about DSI is the focus on using big data and techniques to develop solutions to problems we see in the world. When I was at DSI, I met students who came from different fields of expertise in health and education. I also came across people trying to tackle problems that have a real social impact, which I think is a bit of a different focus then what we see in the usual tech-based programs. I think the fact that more and more trained people bring some substantive experience and deep knowledge in fields like geography and chemistry is very unique.
Can you discuss your work at Mount Sinai?
At Mount Sinai, as a team we are seeing different kinds of tools to deal with quantity, volume or the frequency with which data is collected and these tools are critical for global health in terms of global-health surveillance for emerging infections. We are doing a lot of work trying to estimate disease rates in different places using big-data methods and satellite-imagery analytics. We also try to do things like identify illegal mines or places that are at high risk for large-scale contamination.
Do you use the skills you learned at DSI for your work?
In my coursework at DSI, I was exposed to a different perspective on doing data analysis. I learnedabout problems and developed solutions to them using very large data. I used machine learning techniques that I had never worked on before and that were not commonly used in public health, epidemiology and global health but had proven extremely useful in other fields. So that transfer of knowledge from data science learning to my professional workhas been very useful. Dual training in both DSI and epidemiology is not really common but is something that will continue to become the norm for a lot of people.
When did you first think to combine epidemiology and data science for a career?
The idea of merging data science and epidemiology clicked for me when I was at DSI. My background has always been in public health but in my DSI classes there was a whole new world of modelling and approaches, deep learning and more prediction-oriented work that I had no clue about. Before I came to DSI, I had done a lot of coursework in data visualization and how to communicate complex statistical models for policy makers and different kinds of audiences, not just technical audiences.
Do you feel your work is having an impact?
At Mount Sinai, there’s a lot of emphasis in using data science, especially machine learning, to do better research. Most of that has institutionally focused on medical research and operational work. My team at the hospital is doing something different, which is to apply data science methods to global health. And from what I have seen there has been very little work at that intersection. I’mreally happy doing the work I am doing because I can see a need I am trying to fill. I think it’s been really wonderful to solve legitimate tractable problems. For example, to use data tools and algorithms to help people do better at their respective jobs or for example use tools to help identify kids who are at risk of contracting terrible diseases. It’s very satisfying in terms of what that means for future.
— Robert Florida