Professor Samuel Scarpino has returned to Northeastern University from a year as Vice President, Pathogen Surveillance at the Rockefeller Foundation, with a mission to strengthen Northeastern’s position as a global leader in artificial intelligence and life sciences.
Scarpino has been the director of AI plus life sciences at the Institute for Experiential AI in Northeastern since November. He was previously an assistant professor at the Network Science Institute.
Scarpino recently sat down with [email protected] to answer questions about what makes Northeastern a formidable presence in the AI space. His comments have been edited for clarity and brevity.
What is it about Northeastern that makes the university a leader in AI?
Northeastern’s real strength lies in the importance it attaches to transdisciplinary collaboration and partnerships between business and science.
One reason Northeastern is uniquely positioned to be a leader in this space is because of the word “experimental,” which means cooperatives, but also means “people in the know.” We can do so much more with AI plus humans than with AI alone or humans alone. The idea is that the AI will pick up on things that we might overlook. Humans can also pick up on things that the AI might miss.
Can you give us an example?
The UK has procedures where mammograms are scanned by AI systems. Some of these mammograms are also reviewed by expert panels.
A computer can read 10,000 mammograms without pausing. I can’t remember how many an experienced pathologist can read before having to take a break, but it’s not very many, is it?
Former British Prime Minister Theresa May said AI could reduce cancer deaths by 10% annually.
AI isn’t a silver bullet, but I would definitely subscribe to prevent 10% of cancer deaths. For the final clinical diagnosis, the clinician would have both the AI value and possibly the input of the diagnostic expert panel.
They spoke about the importance of AI in predicting disease states and the tragedy of the existence of anonymized health data that is not used clinically. How could it be used better? And what are the privacy implications?
I’m not an expert on data protection. One of the reasons I joined the Institute for Experiential AI at Northeastern is that it has privacy and ethics experts.
The Institute’s multidisciplinary approach is a big reason I expect Northeastern to be at the forefront of AI. You don’t want to end up with data you can’t use or run into a bunch of regulatory or ethical difficulties.
Northeastern understood that this was an interdisciplinary problem and solved it from the start.
An example of the problems with unusable data is a household study by the Centers for Disease Control and Prevention on whooping cough that began in 2015 and ran for several years. The nasal swab results were never sequenced.
These bacteria could contain evidence of asymptomatic transmission of whooping cough. Sitting in minus 80 degree freezers in Atlanta is very likely the answer to one of the biggest unanswered questions about one of the biggest child-killing diseases left on the planet. And they can never be used unless someone goes back to those 4,000 households and gets permission to use the data.
They do not want data to be exposed and misused. Nor do they want it to be misused in the sense that the cost, effort, energy, and invasion of privacy — which could include blood draws and spinal taps — that went into collecting the data means they’re not being used for anything in the can be used in the future.
dr Naveen Rao, senior vice president of the Rockefeller Foundation’s health initiative, said the foundation is “grateful for the passion, leadership and innovation” you have brought to its pandemic prevention initiative and looks forward to “continuing our partnership to advance pandemic prevention.” strengthen and respond across the global health ecosystem.” How important is the global reach of the Northeast campuses to your work?
A big part of what Northeastern will do is generate network effects, leverage Boston, leverage the Roux Institute in Portland, Maine, London and the West Coast of the United States.
When I work on data for citizens of the European Union, I often have to store and process this data in the EU. And because of our nature of having a broad global footprint, we have the opportunity to work with organizations and on datasets that we couldn’t easily do if we were solely based in Boston.
At the Rockefeller Foundation, my team focused on developing technology products that bridge the gap between data and action. For example, in 2021 we have developed a risk assessment tool for holiday meetings.
We’ve also done some wastewater monitoring, including a joint partnership with NASA and Emory Ghana to design monitoring environments in non-ducted environments.
Which projects are at the top of your list?
One of the first things I want to do at Northeastern is to help build a research data commons where datasets are accessible according to agreed rules on how to use the data.
You would register with a portal and have to agree to privacy restrictions and ethical restrictions according to their terms of service. All of these things are effectively done in advance, meaning that rather than spending three to six months negotiating with the lawyers about what we can and cannot do, this is preconfigured as part of your participation in the research data from Commons.
There’s all these siled datasets in health and life sciences sitting in universities right now. This is an organized way of connecting them.
For media inquiriesplease contact [email protected].
