One of the themes of this year's TEDMED conference was data liberation, kicked off by two speakers: Deborah Estrin, a computer scientist at Cornell Tech and Amy Abernethy, a cancer researcher at Duke Clinical Research Center. Estrin and Abernethy each talked about a rich source of data that could revolutionize healthcare, but isn't currently being tapped; Estrin talked about the "small data" generated by our use of digital devices and Abernethy explored the potential of "data donation."
Deborah Estrin, one of the founders of Open mHealth, spoke at TEDMED not just about mobile health applications, but about how all mobile applications can serve healthcare, by generating what she called "small data."
"We do have special devices designed for self-tracking, like our Fitbits," Estrin said in her talk. "I'm talking about a wider array of data we generate implicitly. We're continuously generating digital breadcrumbs in the services we interact with."
Estrin described how digital devices -- not just smartphones but websites, feature phones, and television cable boxes -- already track data about how much we use them and report it to companies, who use it to optimize advertising and consumer engagement.
"They capture and analyze data extensively, but they don't bring that data back to the person who generated it," she said. "There's no way for us to gain access to that data. There should be, because there's a lot that I can learn about my personal health from my digital behavior."
Estrin acknowledged the concept is similar to Ginger.io's passive monitoring software and that company's idea of a "check engine" light for your health. Estrin said she prefers to think of it as a "social pulse," which can reveal more information about us than a single visit to the doctor -- for which many people will put their best foot forward, rather than acting typically.
"When I think back to my father's last few years of life, I can identify signals that would have shown up in his small data, in his social pulse," she said. "He stopped checking email. His patterns around the neighborhood changed. These things didn't show up in his regular visits to the cardiologist. The attending doctor didn't see anything atypical. There was nothing in his vital signs or his electronic health record, but it would have shown up in his social pulse."
Abernethy talked about a different untapped data source -- individual patients' data. For example, a memorable patient of hers wanted to donate a piece of her tumor to help map out genetic predispositions to melanoma. Abernethy spoke about the need for a data donation infrastructure, perhaps partly modeled after the successful infrastructure we have now for blood donation.
"The blood bank makes great use of every component of my blood. When I need blood, I trust there is a system that will come to my rescue. And when a lot of blood is needed, we know how to band together and donate en masse," she said. "We understand there's a societal institution that supports the donation of a precious resource for universal good."
And there's one big difference between data and blood, when it comes to donation.
"Data is a non-depletable resource, and the more we make use of it, the more valuable it becomes," Abernethy said. "We need to give this data a second life."