The answers to some of today's most pressing biomedical questions may be hiding in medical centers - and physicians' offices - across the country. Buried in medical files are the experiences of thousands of patients, far more than have participated in any clinical trial. These files chronicle their conditions, treatments and outcomes - valuable information that could improve care for millions of current and future patients.
But, and a big but here, accessing the data securely and transforming it into a format available for inquiry can be a logistical nightmare. And that's where Stanford's Daniel Rubin, MD, and his team, step in.
"We're developing methods that will permit us to leverage national data without requiring centers to actually send the data to a central site, which overcomes a big barrier to these kinds of efforts because of privacy and other regulatory concerns," Rubin said in the video above.
His group is concentrating on cancer and received a 2014 Big Data for Human Health Seed Grant to support the work.
"We've developed software that we've deployed at these local sites... They run it on their local data and we aggregate the results," Rubin said. He said the primary challenge is creating a system that is open and invites broad participation, but also keeps the data secure. His project exemplifies the uses of big data. From the video:
You couldn't do this without big data because there are so many variables that affect a patients' disease... and you need big data to find enoguht patietns that match those charactoristics to be able to look for similar cohorts to guide decision making.
This effort is part of Stanford Medicine's Biomedical Data Science Initiative (BDSI), which strives to make powerful transformations in human health and scientific discovery by fostering innovative collaborations among medical researchers, computer scientists, statisticians and physicians. For more on important work being done in this area, mark your calendars for Stanford's Big Data in Biomedicine conference on May 20-22. More information is available here.
Previously: All data - big and small - informs large-scale neuroscience project, Examining the potential of big data to transform health care, Registration for Big Data in Biomedicine conference now open and Stanford researchers develop web-based tool to streamline interpretation of medical images,