Fighting TB with technology
- Published on 13 July 2012
Two South African students have developed software that could help medical specialists in the country detect the lung disease Tuberculosis (TB) faster.
Joshua Leibstein and Michael Cilliers, from the University of Johannesburg, have created pattern-analysis software that they say can help detect the highly contagious disease in scanned radiograph images by up to 20 minutes faster than the traditional process.
The software, in conjunction with either cloud computing or graphics processing unit (GPU) technology, uses a general textual analysis algorithm that identifies instances of TB in digital scans of patients’ radiographs.
They showcased their tech at the Microsoft Imagine Cup Finals 2012, which was held in Sydney, Australia this past week. The competition aims to let students develop solutions to the world’s most pressing problems, and winners in the finals of the Imagine Cup also receive funding to help take their innovations to market.
However, although the UJ students impressed the judges, they didn’t make it through to the second round of the competition. But the young men say that there is a future for their software, in a country where the disease rates are up to 60 times higher than in the US or Europe.
“It’s going to continue,” says Leibstein, speaking about the project’s development.
“The actual system would not be expensive, because it’s not using any specialised hardware,” he added.
The two students say that they are also in talks with Johannesburg’s Helen Joseph Hospital to roll-out the tech for the facility.