Machine learning set to transform the face of global healthcare
Revolutionizing healthcare the intersection of computer science, big data, and life sciences. Leading the revolution in disease prevention, detection, and treatment for humanity.
Anantha P. Chandrakasan, Chair of J-Clinic, talks to Opening Doors about his hopes for the newest MIT/Community Jameel partnership
Anantha P. Chandrakasan, Dean of the MIT School of Engineering and the Vannevar Bush Professor of Electrical Engineering and Computer Science, is the recently appointed Chair of the Abdul Latif Jameel Clinic for Machine Learning in Health (J-Clinic). Over the next few years, Anantha and the J-Clinic team will be pushing forward the boundaries of global healthcare research to deliver real-world impacts in disease prevention, diagnostics, and drug discovery and development.
Opening Doors spoke to Anantha about the latest partnership between MIT and Community Jameel, and how he plans to maximize its impact in the coming years.
Q: When you first heard about the proposals for J-Clinic, what was your initial reaction?
AC: I first discussed it with Mohammed Jameel in September 2017. Straightaway, I was excited by the opportunities of combining global healthcare with some of the leading-edge AI research at MIT to enable new ideas to emerge.
We already had people working on machine learning at MIT, and others on healthcare, and there was some work at the intersection of the two. I saw J-Clinic as an opportunity to bring that intersection together and make a truly global impact with our work in this crucial field.
Q: How heavily involved was MIT in these fields before the creation of J-Clinic?
AC: We were already involved in research around AI and healthcare, linked to our proximity to Boston’s world-leading healthcare cluster. We have the Institute for Medical Engineering and Science (IMES) here at MIT, the Computer Science and Artificial Intelligence Laboratory (CSAIL), and many more departments, centers and labs at MIT. In fact, when we were defining our ‘Quest for Intelligence’ initiative, we found there were 200-plus principal investigators at MIT working on AI related topics, including healthcare-related research.
J-Clinic is a way to encourage collaboration between them all, because the innovation we’re seeking is only going to happen if we bring together researchers working on machine learning and AI with those working on healthcare technologies.
We also need to foster close connections with the hospitals and with the industry players and innovators that can help to commercialize some of these ideas. Collaboration and innovation have little value in isolation. There is a whole ecosystem that needs to come together to make sure innovation is translated into impact, and that’s what makes J-Clinic so exciting. J-Clinic can be the catalyst for this new type of collaboration.
Q: What has been the current impact of AI in healthcare, and how much further is there to go?
AC: It’s still in the very early stages. The idea of AI, and particularly machine learning, has recently re-emerged because of advances in image recognition technology, the exponential growth in the power of computing/GPUs, and the availability of huge amounts of data.
Early breast cancer detection has been one of the successes so far – in fact, one of our faculty colleagues, Regina Barzilay, has done some pioneering work on the use of machine learning for breast cancer detection. But overall, AI in healthcare is still a very new field.
What really excites me about this whole area is the potential for these new technologies to have an impact far beyond big hospitals in big towns and cities. It could transform healthcare in a wide range of settings worldwide, including rural facilities and emerging markets.
Q: Are we talking about a fundamental revolution in healthcare, or is it just a case of improving current practices?
AC: A bit of both. Certainly, it will streamline and make much more personalized healthcare possible. We will be able to track an individual’s data, understand it in relation to the wider population, and customize treatments more individually. That’s a very exciting opportunity.
But ultimately, I’d love to see this technology looking at the prevention of diseases. These are the more futuristic opportunities. Perhaps we’ll be able to detect tumors even before radiologists can see them on a mammogram – that’s the kind of ambition we’re working towards. The first achievement is likely to be detection, then better treatment, then personalized treatment, and finally disease prevention. That’s our ultimate goal.
Q: Using AI and machine learning for disease prevention sounds like science fiction. Is it realistic?
AC: Not for every disease, but I believe we can certainly make progress on several different fronts, enabled by the kind of unique collaboration that J-Clinic can provide.
One of the biggest challenges is around access to data, because today very little of the data that’s collected is actually used. There is bound to be hidden knowledge buried in that data, we just can’t access it. In time, I believe there will be advances that allow us to access a lot more of the data that’s collected.
Q: How quickly might we see some practical results from J-Clinic?
AC: It’s hard to say. There have been a lot of innovations recently, but there need to be many more to get to where we want to be. So it will be a combination of both long term and short term results. This is reflected in the way J-Clinic is set up. Some of the funding is going to target the long-term, more basic research that has a longer time horizon. Other funding will target solutions that are closer to commercialization and look at how we can help to accelerate these ideas from the lab to impact, working with MIT’s Deshpande Center for Technological Innovation.
Q: How did you decide on J-Clinic’s three stated areas of focus – preventative medicine, diagnostic tests, and drug discovery and development?
AC: We looked to see where we could make the biggest impact. With diagnosis, for example, I’m thrilled that J-Clinic will not just target the hospitals with the best facilities. We want to do diagnosis in rural settings, like India, for example, where they may not have access to mammograms or there may not be a single doctor for an entire village. The promise of bringing AI technology to healthcare environments that may not have the kind of resources enjoyed by the developed world is incredibly exciting.
In preventative and more personalized medicine, it’s exciting to consider being able to target and create drugs for a particular person. At the moment, it’s a completely open playing field. And we can also use machine learning to guide the process of drug discovery, creating drugs that target very specifically-defined groups of people.
Q: What are your ambitions for J-Clinic over the next five or 10 years?
AC: In the short term, we’ll look to develop new algorithms and have an immediate impact with some of our ongoing work on detection, diagnosis, and drug discovery. I’d like to explore how we can make partnerships with both hospitals and industry, particularly with regards to deploying our ideas, and we also want to spin out our research and create start-ups to help drive the next generation of healthcare innovation.
As a personal goal, I’d like to take this technology to all parts of the world. Our researchers love to solve these kinds of problems, and I’m very keen to encourage them to find the solutions needed to move this technology forward worldwide.
Set against those ambitions, though, is the reality that getting access to data will be our biggest challenge. It is something we will have to solve. The better access to data we have, the better our prediction and treatment approaches are going to be.
Finally, some of our research has to be focused on the ethical use of these technologies. Those questions are not scientific, but they are equally important if we want to successfully deploy our research.
Q: Could AI and machine learning close the gaps between the quality of healthcare people can access in different parts of the world?
AC: Yes, I believe so. That’s a big part of it. You can certainly envisage a situation where machine learning acts as the first step in deciding who, in a rural community, should go to a bigger hospital in the nearby town, for example. But there is a lot to do before that. What new kinds of devices will we need to take these new systems out into the real world? What innovations and algorithms will we need to develop to make best use of the data we have? And can we work with governments to subsidize some of the equipment healthcare workers will need?
Q: How pleased are you with the strength of J-Clinic’s leadership team?
AC: I’m thrilled. Having two MacArthur fellows as the co-leads, covering machine learning and health from different perspectives, is something to be very excited about. Professor James J. Collins is the leader of synthetic biology, and Professor Regina Barizlay, is the leader of machine learning for health. Then there’s Institute Professor Phillip A. Sharp, a Nobel laureate who has advanced basic science and brought his innovations into the commercial world, giving us expertise in research and entrepreneurship. All these people are experts in their field, and all of them also have extensive experience working with the healthcare industry itself.
Q: Finally, J-Clinic is the fourth MIT collaboration with Community Jameel. How important has the MIT/Community Jameel partnership been over the last few years?
AC: It’s been absolutely instrumental. Our students want to make an impact on the world. So when you look at the Poverty Action Lab (J-PAL), the Water and Food Systems Lab (J-WAFS) and the World Education Lab (J-WEL), these are the kinds of deep and important problems our students want to address through their research. J-Clinic fits perfectly into this environment.
These programs have been absolutely game-changing. I’m particularly excited about the opportunity for the programs to interconnect and amplify each other’s impact over the coming years. It’s a great opportunity for J-Clinic. We hope to partner with the other J-Labs in very specific ways in both the short and long term.
The partnership with Community Jameel has provided us with much-needed resources to facilitate this valuable work. But more importantly, it’s enabled us to develop a research framework in which our students thrive, creating connectivity that can help to make a practical, real-world impact.