Computer Program to Assess Diabetes Risk with Increased Accuracy
Newswise — An innovative computer program designed to help health professionals identify the risk of developing Type 2 diabetes is being prototyped at Trinity Western University (TWU), thanks to seed money from a two million dollar donation.
While there is currently no cure for the disease that afflicts over two million Canadians, Type 2 diabetes—the more prevalent of the two types—is preventable. TWU computing science professor, Alma Barranco-Mendoza, PhD, is leading one of the University's three, 3-year studies established to further diabetes research and to ultimately find a cure.
The TWU Diabetes Research Endowment Fund has already yielded promising results. An expert in diagnostic systems, Barranco-Mendoza is developing a revolutionary intelligent system able to assess diabetes risk with increased accuracy.
"There are effective diagnostic tools available to doctors and nurses," explains Barranco-Mendoza who has already landed nominations for the 2004 BC Systems Institute Innovations Award and the 2005 NSERC PhD Dissertation Award, "but none of them employ intelligent profiling. Their capabilities are limited to a general estimate of risk since they are only able to take one risk factor into account."
Thanks to her background in cancer diagnostics, Barranco-Mendoza has successfully developed a sophisticated algorithm capable of assessing and weighing numerous risk factors—including lifestyle, race, genetic information, family medical history and pancreas tissue images. This algorithm sets Barranco-Mendoza's computer program on course to become the first multidisciplinary diagnostic tool for diabetes. Instead of delivering a general reading of low, medium or high—what patients currently receive—the new program will weigh the various factors and express risk as a specified percentage.
"Dr. Barranco-Mendoza developed a unique, highly original algorithm that represents an outstanding contribution to the field of Computing Sciences," says Veronica Dahl, PhD, Professor of Computing Science at Simon Fraser University and Barranco-Mendoza's PhD Supervisor. "It rests on state-of-the-art developments in logic programming, but it goes far beyond. By incorporating complex theories, the program is able to simultaneously consult and draw inferences from diverse and even incomplete sources."
And Barranco-Mendoza is not only creating a computer prototype, she's making headway in diabetes research data collection, representation and analysis, collaborating with the BC Genome Sciences Centre and the Medical Image Analysis and the Diabetes Research Laboratories at SFU. Together these groups can develop more comprehensive profiles of "at-risk" patients, thereby increasing the understanding and precision of diabetes risk-assessment in general.
"Once a certain amount of data is gathered," says Barranco-Mendoza, "the intelligent system will start learning by itself. It would mine the data and identify trends not evident to us. It is believed that there may be DNA markers for the disease and if we can identify them we're on our way to finding a very fast and effective early diagnostic test and even possibly making headways to eventually find a cure."
Barranco-Mendoza's algorithm development is one of three diabetes research projects underway at TWU. The second study partners TWU's nursing and kinesiology departments with the SeaBird Island First Nations community in order to explore strategies to identify and reduce risk factors for Type 2 diabetes in aboriginal children and youth. The third project involves TWU biologists who are investigating the disruption of insulin signalling using powerful genetic, proteomic and cell biology tools to examine specific cell behaviours that regulate glucose metabolism in fruit flies and model worms.
Print-friendly Layout