A team of Dartmouth researchers led by Cory Cornelius '07, who is currently pursuing a PhD in computer science at the College, published a paper this month detailing how a sensor bracelet can use biometrics to passively identify individuals for health care purposes. The paper, titled "Who Wears Me? Bioimpedance as a Passive Biometric," was overseen by computer science professor David Kotz and is a part of the Institute for Security, Technology and Society's research initiative on information systems and health care. Cornelius, Kotz and other researchers involved in the project emphasized that the research is still in its early stages and they have several issues to address before the sensors become a technology that could gain widespread traction. The sensor bracelet is designed to authenticate the wearer and detect a range of medical needs, both for convenience and for potentially aiding in life-threatening situations, according to Jacob Sorber, a post-doctoratal student who advises Cornelius' research.
The device is also compatible with fitness applications on computers or smartphones to differentiate between users sharing the device. A watch that collects heart rate information, for example, is unable to distinguish between various members of a family, whose information will be jumbled together.
Smoking cessation programs could also use the sensor bracelet when collecting data with a device to ensure compliance, Sorber said. The bracelet would verify that the right person is wearing the compliance sensor, as the smoker could be tempted to cheat by asking a non-smoker to wear the compliance device. "The general approach is to provide confidence to the data coming in," Sorber said.
Rianna Starheim '14, the only undergraduate on the research team, said that many medical devices require authentication, including insulin injectors for diabetics.
"You don't want a kid getting ahold of that device and accidently injecting himself," Starheim said. "It's vital to have some kind of authentication process to ensure that a user is the intended user before a device carries out its function."
The sensor bracelet uses bioimpedance a measure of "how the body's tissues oppose a tiny applied alternating current" to identify individuals, according to the paper. The distribution of bone, fat, muscle and other tissue types in each person's body is unique and can be used for identification.
While most biometrics, such as fingerprints and voice authenticators, require action to work, Cornelius said that one of his goals was to create a device that would provide personal authentication without significant effort from the user.
"We wanted something that would integrate into a person's life seamlessly and passively," Cornelius said.
In the fall, the research team plans to tackle challenges related to the stability of bioimpedance as a metric, according to Starheim. Proportions of fat, muscle, water content or sweat may change in a person's body over time. If bioimpedance varies too much in one individual, the variation between people is no longer relevant, Starheim said.
Some measures of bioimpedance are more variable than others, and the project will have to find those measures that are less changeable and use them, Sorber said.
"We need to look at how these impedance measures change over time," Sorber said. "Some of these measurements will be more stable."
Cornelius pointed out that the location of the sensor bracelet mitigates problems of instability because fat and muscle are not gained as significantly on a person's wrist.
He said he has no aspirations to commercialize the device and is focused instead on completing his PhD by 2013. In addition, members of the team said that the technology requires much more research before they can validate the device's effectiveness.
The research project is supported by a grant from the National Science Foundation, which was provided to ISTS for the Trustworthy Information System for Health Care effort, according to Kotz.
TISH is an initiative that seeks to innovate information-sharing technology in health care to increase security and privacy, according to the website. Various threads of research within the TISH initiative include mobile health care, which is spearheaded by Kotz and includes Cornelius' project.
"We are trying to address the fact that health care is supported by information technology," Kotz said. "If these devices aren't well-designed, they may share information of a personal nature you don't want to share."
