CMLH’s Vision for 2021

We had a chance to sit down with CMU’s Center for Machine Learning and Health Executive Director, Dr. Joe Marks, and Chief Science Officer, Dr. Carl Kingsford, to talk about their vision for both the CMLH and PHDA in the year ahead.

Thinking back to 2020, what do you think the PHDA’s greatest accomplishments were?

Over the last few years, we have created an innovation community in digital health between UPMC, CMU, and Pitt. Innovation is complex and it boils down to community – people sharing knowledge, learning from each other, and helping one another. This community that we’ve helped to create has weathered 2020 and will rebound in 2021. Additionally, in 2020, the CMLH Fellows did a lot of interesting foundational work in digital health and we also continued to forge our relationship and work with AWS.

What were some of the greatest challenges the PHDA faced in 2020?

It’s undeniable that 2020 was a challenging year for everyone. Maintaining morale was critical and fortunately, we tried our best to focus on the light at the end of the tunnel. The essential parts of what has made the PHDA successful to this point are still in place and we are ready to rebound and learn. We continued to support the startups that we have fostered within our community despite the obvious challenges that they were faced with and did so while maintaining morale.

What are some current challenges and opportunities in the digital healthcare space and how do you think the Center(s) can address them in the future?

There are many places where the needs in digital health intersect with the interests of CMU’s researchers, and over the course of 2020 we deepened our understanding in terms of the breadth of available research talent and interest at CMU. Some opportunities include machine learning techniques for histopathology, personalized medicine, home healthcare, biomarker discovery, device manufacturing, and measuring responses to drugs. There are also the less obvious areas such as hospital operations and efficiencies, billing, and logistics, which match with CMU’s expertise.

We also did a census of CMU faculty and their research at CMU that is relevant to digital health. What we found was that we knew, and were working with a lot of these people, but there were also many groups and individuals that we did not know about. The exciting thing about the future is that we have the opportunity to now engage with these new people and share resources in order to continue to deepen our impact in digital health.

How do you envision furthering your collaboration efforts with your colleagues at CMU, Pitt, and UPMC?

We hope to continue to expand our proactiveness in reaching out to the people within CMU’s School of Computer Science and getting more engaged in digital health through the PHDA. Virtual outreach will help us broaden the community of people that are working on these efforts. There’s always been a challenge, and an opportunity, around data sharing and data sets. A big part of the AI revolution is machine learning, and there’s no machine learning without data. Some of the greatest successes in terms of projects and spinouts that we have had have been due to the access that they had to unique data sets as a result of our collaboration with UPMC and Pitt.

Have there been lessons learned by the Center because of the pandemic? If so, what and how has it changed things?

We always knew that telemedicine was going to be important, but 2020 made it extremely obvious that it’s a critical area where we can continue to focus on research. Also, virus-based research is something of renewed interest. The pandemic also brought social determinants of health (SDoH) and population health into focus, specifically as it relates to healthcare and the uneven impact that the pandemic has had. We would have liked to fund more research and get more people involved in projects that specifically relate to the behavioral and economic aspects of health. Moving forward, we’d like to see an increased focus on this type of work.