The applications for data science in healthcare are far-reaching. The ability to predict potential health issues is, obviously, hugely beneficial as catching illnesses early greatly improves the chances of it will be possible to cure them. Data science can also be used for things like improving chronic disease management initiatives, lowering hospital readmission rates, and tracking the spread of contagious diseases. The move to a predictive model of healthcare should, and already is, greatly reducing costs in an industry currently under immense pressure.
Eileen B. Leary is Senior Manager of Clinical Research at Stanford University. She began her career more than 20 years ago as a sleep technologist before transitioning to data science and obtaining her Masters in Epidemiology and Clinical Research from Stanford. As the Senior Manager of Clinical Research at the Stanford Center for Sleep Sciences and Medicine, Eileen is directing a massive project to improve the understanding sleep and sleep disorders by creating a large cohort and sharing it with the scientific community. In her spare time, Eileen is pursuing her PhD at Stanford to unravel the mysteries of sleep.
We sat down with her ahead of her presentation at the Big Data Innovation Summit, which takes place in San Francisco this April 12 & 13.
How have you seen the data landscape change in the past 5 years?
I’m a data nerd, so I’m thrilled that so many scientific journals have started implementing data sharing policies. Authors of accepted papers in many journals have to share whatever data and code are needed for replication. This required level of transparency will lead to improved science and data management practices. It also magnifies the contribution of each individual research subject by expanding opportunities to reuse datasets.
Which developing data technology are you most excited about?
I’m really excited about the growth of health sensors that allow passive collection of physiological signals and important health metrics. So many of us are disconnected from what’s happening with our health, which is a huge impediment for reporting subjective symptoms. The ability to accurately collect biological data outside of a clinical environment for an extended period of time will allow for early diagnosis and intervention for all types of medical problems, ranging from cardiovascular health to sleep medicine.
What are the biggest challenges that the data function currently faces?
Data sharing comes with a number of significant hurdles. Many scientists don’t have the knowledge or experience to implement optimal data management and documentation practices. They use naming conventions and coding systems that are not intuitive and rarely bother with a comprehensive data dictionary. In addition, as the size of datasets continues to grow, it is more cumbersome to securely share large datasets across institutions for analysis.
How do you see data changing in the next 5 years?
I’m not sure that medical data itself will change significantly over the next 5 years. But we will certainly have a lot more of it. We will also have new tools and techniques to both collect and analyze the information.
What can the audience expect to take away from your presentation in San Francisco?
My talk will focus on how signal processing and machine learning are being used to automate the analysis of sleep data and improve our understanding of different sleep disorders. By moving beyond manual scoring and the obvious biomarkers being used today, we hope to create new diagnostic tools. Of course, the ultimate goal is to enhance treatment options for sleep disorders.
You can hear more from Eileen, as well as other industry leading experts in the data field, at the Big Data Innovation Summit. View the full agenda here
BONUS CONTENT: Watch Rajiv Bhan, Manager of Data Science and Engineering at Fitbit, discuss how the wearable giant used machine learning to personalize the consumer experience