Cancer researchers pioneer data mining techniques

Medical researchers at the University of Leeds are pioneering the latest data mining techniques in their fight against Leukaemia.

The Disease Registry Application (DRA) is designed to allow researchers to map, analyse and display data on the distribution of the disease nationwide.

The software is the result of a three-year collaboration between Leeds University's Computing Services, the Leeds Research Fund Centre for Epidemiology and data warehousing company, SAS Institute.

Paul Nicholson, head of application software, Leeds University, said: "One of the major benefits is that we now have an application which integrates most of our analysis into a single framework, and this has given a significant increase in productivity for the end user.

"Another important aspect of the application is that it serves as a protocol for doing this type of work and guarantees that the results obtained by different workers - either in the same team or different teams - are consistent."

Bruce Bouvill, SAS project manager for the educational market, said the application design had been lead by the development team at Leeds. "They built a data warehouse so they could take different types of source data and bring them together to be normalised for age group, ethnical background etc.

The Leeds team became interested in using the software after noticing the similarity between business organisational structures and the hierarchical data in the disease registry.

"It's considerably easier to do the work now," said Nicholson. "Researchers can carry out analysis in seconds rather than hours. And they can have much more confidence in the data."

The work of the development team at Leeds was recognised today when they also received an award from SAS for excellence in academic computing. The application is now being trialled in hospitals in Manchester and Leeds to map occurrences of childhood cancer and diabetes.