Healthy Tech Ventures



Dr. Nathaniel Osgood

Hand-held research labs

Over in the Department of Computer Science, what began as a smartphone app has morphed into an enterprise that is taking health-care research in directions hard to imagine a decade ago.About eight years ago, associate professors Nathaniel Osgood and Kevin Stanley began work on a sensor device to track the movement of study participants, useful data for predicting health risks such as how flu viruses may spread. The project developed into the iEpi smartphone app. They realized that mobile devices “can be formidable for not only measuring what (people are) doing at a given time, but also for asking them questions and for getting them to volunteer information, let’s say when they’re taking some medication … or feeling ill … or if they’ve just encountered an advertisement about tobacco or what have you,” Osgood said.

The app collects information from sensors built into the smartphones, along with information volunteered through questionnaires, photographs, even sound and video recordings. It can be linked with wearable devices such as Fitbit and other smartwatches to measure heart rate and electrodermal response, an indicator of stress or arousal. It is a micro research lab that enables the collection of an unprecedented level of data, both in volume and detail. For instance, when a study participant takes a photo of a tobacco billboard, researchers will know where it is, what time the person came across it and if they smoked afterwards.

Before long, Osgood and Stanley were approached by researchers wanting to use the app in some very large studies, with tens or hundreds of thousands of participants. So, last year with help from the Innovation Enterprise office at the U of S (formerly the Industry Liaison Office), they formed a start-up called Ethica Data. The company collects data from smartphones (either Androids or iPhones), aggregates it, streams it to servers, and allows both study administrators and subjects to browse their own data. “What we’ve designed is a robust, easy-to-use, flexible, powerful and integrated system,” Osgood explained.

Notably, it enables health researchers to very quickly and easily launch a new study without the need for custom programming and the help of computer scientists, Osgood said. Think of on-the-spot research into sudden public health crises such as food contamination or disease outbreaks. But the system is also valuable in researching such intractable problems as tobacco addiction or the high rate of suicide in northern communities. Or, studies to answer such questions as whether e-cigarettes help break nicotine addiction or act as a gateway. Another key feature: the capacity to not only obtain informed consent from study participants online, but also to enable them to suspend data collection at any time, even delete their data retroactively.

Osgood maintains this does not jeopardize the validity of a study, because in practice very few people exercise this option. And yet it’s important, particularly to disadvantaged people, who may be anxious about providing personal information. And in a kind of spy vs. spy scenario, Ethica can also capture the internet browsing patterns of study participants (with their consent). For instance, when they click on links with tobacco-related words researchers can learn about the advertising they see. Osgood said corporate interests are also mining data from online behaviour and using it to promote their products. “And by providing the option of people to join studies like this we provide an avenue for us to counter that effectively and in a very flexible way,” he explained.

Osgood and others, including various public health agencies, are now looking at creating an ongoing pool of respondents they can tap into for studies from time to time, a kind of population observatory or sentinel group comprising perhaps five per cent of the population. This proverbial canary in the coal mine can provide an early warning of threats, and allow for both quick responses in an emergency and better decision-making overall, Osgood said. Perhaps some data will be held in escrow, he goes on, to be retroactively mined for retrospective analysis. And, using big data or machine learning techniques, we may one day see the creation of digital nudges that prompt people to, say, dial up the tobacco quit line when they pick up a cigarette. “I think here we have a wide-open set of possibilities and it’s a privilege to be at the forefront of this,” Osgood said. 

Full article, plus more health tech ventures at the University of Saskatchewan