Features
CIS Research Aims to Help Computer Users, Doctors
BY ANDREW VOWLES
Desk-bound workers often blame repetitive stress injuries on their computer keyboards. Now U of G researchers hope to use computing smarts to help those workers get better faster and — in the process — help improve relations between people and their silicon-based deskmates.
Helping to make a more user-friendly computer is what drives research by Tyler Doan, a master's student in the Department of Computing and Information Science (CIS). Along with his co-supervisor — adjunct professor Andrew Hamilton-Wright — and Robert Varga, another CIS master's student, he also hopes to help doctors better pinpoint and treat hard-to-diagnose disorders affecting muscles and nerves.
The group aims to develop software that will churn through patient data and help a doctor analyze and diagnose a problem. In effect, their decision-making tool in the doctor's office or clinic would become the MD's electronic helpmate in assessing patient health.
“It's all about helping clinicians understand the information they've acquired,” says Hamilton-Wright, who is using data on carpal tunnel syndrome to test the system.
Those spiky graphs on his laptop show patterns of electrical signals recorded from muscle nerves of patients suffering from carpal tunnel — or not. That ambiguity is partly the point, says Hamilton-Wright, a two-time Guelph graduate who is now a faculty member in mathematics and computer science at New Brunswick's Mount Allison University.
By helping doctors cut through the “noise” around disease symptoms, he hopes to help them diagnose and treat this progressively painful condition often encountered at work when repetitive motions cause nerve compression in the wrist.
He's studying data collected from several clinics. One source was the School of Rehabilitative Therapy at Queen's University, where he spent a year working with clinicians on patient data sets.
Beyond carpal tunnel, the researchers plan to develop tools to help doctors better diagnose other disorders. Right now, it's difficult for clinicians to clearly distinguish between various myopathies (muscular dystrophies, polymyositis) and neuropathies such as Guillaume-Barre syndrome and Parkinson's disease, says Hamilton-Wright. Helping to make better diagnoses could improve patient treatment and save money for the health-care system, he says.
Doan began his master's program last fall after completing a B.Sc. in the department. He's co-supervised by Prof. Judi McQuaig, who studies human-computer interactions. (His twin brother, Adam, is studying genetic algorithms used in neural networks with CIS professor Mark Wineberg.)
Broadening the scope, Doan is interested in learning how to make it easier for people to understand data presented on a computer screen.
Looking at a 3-D diagram on his laptop, he says the blue cubes, green cylinders and other shapes represent data from a heart study. He's been recruiting subjects to visit his Reynolds Building office to spend 30 to 60 minutes viewing such pictures of data. “Can people look and tell me what it means?” he says.
Knowing whether people can make sense of what they see may help in improving computing tools and in allowing computers and their users to work together. He's interested in the cognitive rules that underlie decisions, whether made by people or by a computer.
Doan says their work might enter other applications where users need a computer to serve as another “mind” in assessing information and reaching a decision.
“That's the beauty of computer science. You can apply it in a lot of places.”