Xerox fellow analyzes motion data from clinical trial

Xerox fellow Matthew Dombroski is analyzing motion data gleaned from sensing devices worn by participants in a clinical trial, as a way to detect and assess Parkinson’s and Huntington’s disease.

Last year, MC10, Inc., a pioneer in biometric-sensor-enabled analytics, announced an exciting collaboration with the University of Rochester.

The Massachusetts company would unite its biometric sensing devices, companion software application, and end-to-end cloud storage and computing platform with the University’s clinical expertise and big data analytics. Together, they would drive solutions for today’s most pressing healthcare challenges.

This summer, Xerox fellow Matthew Dombroski, a rising senior in electrical and computer engineering, is helping that collaboration by analyzing motion data gleaned from MC10 sensing devices worn by participants in a clinical trial at the University of Rochester Medical Center.

“The hope is that we can detect and assess Parkinson’s and Huntington’s disease,” said Dombroski, who is mentored by Gaurav Sharma, professor of electrical and computer engineering.

  “This is a topic I haven’t worked with before,” Dombroski said. “But it involves something I’m super interested in – machine learning.”

This is interdisciplinary research at its best, combining clinical expertise from the Medical Center and engineering expertise from the River Campus. Sharma said the project is particularly exciting because the sensing device is actually worn on the body, much like a bandage, eliminating the extraneous motions that would be sensed by someone using a handheld sensor, for example.

  “We would like to develop and define the accelerometer profiles as signatures which correspond to the at-rest tremors of individuals who have Parksinson’s or to other motion irregularities,” Sharma said.

He credits Dombroski with augmenting an available app so that it can be used to calculate the duration of each step, which tends to vary with different disease conditions.

  This is the first research experience for Dombroski.  “It’s very nonlinear,” he said. “In other types of work, the more time you put in, the closer you get to your destination. In research, at least in my experience so far, you can spend 10 hours and get nowhere, and then, in one hour, everything can fall into place.”

“It’s interesting,” says Dombroski, who plans to attend graduate school.  ”It’s a new challenge.”