Jessica Mega of Verily Discusses AI in Healthcare

Jessica L. Mega, MD, MPH, is the Chief Medical Officer at Verily Life Sciences. As a faculty member at Harvard Medical School (on leave), a senior investigator with the TIMI Study Group, and a cardiologist at Brigham and Women’s Hospital, she led large, international, randomized trials evaluating novel cardiovascular therapies. She also directed the TIMI Study Group’s Genetics Program. Her research findings have been published in the New England Journal of Medicine, Lancet, JAMA, and elsewhere. Dr. Mega is a graduate of Stanford University, Yale University School of Medicine and Harvard School of Public Health. Read her full bio.

Jessica Mega of Verily Discusses AI in Healthcare

Q: Artificial intelligence (AI) techniques have sent vast waves across healthcare, even fueling an active discussion of whether AI doctors will eventually replace human physicians in the future. Do you believe that human physicians will be replaced by machines in the foreseeable future? What are your thoughts?

A: In addition to the science, the art of medical practice is a critical dimension of patient care and that part of the human experience can’t easily be replaced by AI. Like any other professional group, physicians begin using new technology regularly once it demonstrates added value and, in this case, that it can improve patient care. Adoption of common diagnostic tools like the stethoscope and echocardiograms have proven critical in identifying people at risk of disease so they can receive earlier treatment. In terms of care delivery, machine learning tools that likewise demonstrate they can improve outcomes will be used to augment the practice of medicine.

Q: Can you provide some use cases that have already successfully demonstrate the value of AI/Machine Learning in healthcare?

A: An area where machine learning is already having an impact on care is using retinal imaging to screen for early signs of disease that can cause blindness. Diabetes is a healthcare issue of pandemic proportions, and complications related to diabetes, such as diabetic retinopathy, are one of the leading cause of blindness in adults. Diabetic retinopathy can be treated, however there are barriers to screening given the proportion of ophthalmologists to patients, particularly in countries like India. Google and Verily’s work has shown that automated deep learning algorithms can improve upon manual grading by ophthalmologists to identify diabetic retinopathy1 and we have a partnership with Nikon Optos to integrate cameras with image recognition and develop a medical product that can be deployed by general practitioners, for example, to increase access to screening.

Q: What areas in healthcare will benefit the most from AI/Machine Learning applications and when will that be?

A: Both physicians and patients are constantly inundated with information and challenged to determine what’s critical to decision making. Machines can be better at detecting particular patterns than we are, and so machine learning can help to identify where to focus our efforts in terms of patients at risk of disease, for example. It can also help us to get ahead of poor outcomes by scaling our ability to recommend tailored interventions and strategies for patients.

Q: What are some of the challenges to realize AI/Machine learning in healthcare?

A: Healthcare data is still fragmented and that’s problematic for machine learning, which requires a broad set of data points so you have the information and context need to train generalizable algorithms. We, as a community, need to harmonize how data appears across organizations to support automated ingestion of data. The next challenge, assuming you’ve secured and formatted quality data and created the machine learning model, is to develop that model into an approved medical solution that demonstrates real value. In the world of value-based care that we all have our sights on, that value will be measured in both outcomes and dollars.

Q: How close are we with successfully using AI for the purpose of mining big data?

A: There are examples of robust work that are ongoing. For example, our colleagues at Google are partnering with UCSF, Stanford Medicine and University of Chicago Medicine to explore how EHR data can be used to predict medical events like hospitalization or length of stay, and to anticipate patient needs.

Q: What is your outlook or vision for use of AI/Machine Learning in healthcare?

A: The real challenge in healthcare is supporting patients and clinicians to make the most optimal decisions out in the real world. Machine learning has the potential to have a radical impact when it comes to driving these decisions since it helps us to surface insights from a continuous data stream, and to rapidly determine what interventions or recommendations will derive the desired action from a patient. So, the promise of machine learning is that it can take some of the labor and complexity out of care by providing insights to physicians and precise recommendations to patients.