Image processing and fuzzy logic to automate cancer detection

JULY 14, 2008--Technology that will make it possible to semi-automate the process for detecting cancerous tumors more precisely is being developed by researchers at the University of Hertfordshire (www.herts.ac.uk). Soodamani Ramalingam and her team at the university are developing 3-D object recognition and image processing to make it possible to get a more accurate picture of human tumors so that cancers can be identified and treated accordingly. They plan to make this technology available within three years to hospitals throughout the UK and seek collaborators and further funding to make this happen.

In collaboration with the Paul Strickland Scanner Centre, where clinical image data are acquired, the team is using a combination of fuzzy logic (a type of logic that recognizes more than simple true and false values) and image processing to identify cancer and establish accurately how far it has spread. "This new method of image analysis will be more accurate in defining tumor edges and will potentially allow more effective treatments," said Ramalingam.

According to the researchers, the technology, which will also have applications in other medical fields, will complement traditional positive emission tomography (PET) and computer tomography (CT) scans because the fusion of fuzzy logic and image processing will produce better defined tumor images and so greater certainty in diagnosis and treatment. "The classical mathematical approaches to looking at clinical scans do not always produce high-resolution images," said Ramalingam. "In some cases this means that specialists cannot interpret the PET/CT scan very easily, so there is always a certain degree of uncertainty that our advanced analysis techniques will address."