Important Step in the Early Detection of Colon Cancer

The “Computer Technologies in the Early Diagnosis of Colon Cancer and Brain Mapping” research project developed by Asst. Prof. Burak Acar at Boğaziçi University made significant headway in the early detection of colon cancer by virtual colonoscopy.

Boğaziçi University academicians have developed a computer-assisted diagnosis method for the early detection of colon cancer.Based on processing and analyzing a three-dimensional medical image of the area, the approach is a virtual colonoscopy that is non-invasive, unlike classic colonoscopies.This newly-developed method makes it possible to significantly reduce the margin of error in computer-assisted colic polyp detection algorithms by making colonic polyps—the precursor of colon cancer—more visible on CT scans.

Project coordinator Asst. Prof. Burak Acar of Boğaziçi University’s Department of Electrical and Electronic Engineering said that virtual colonoscopy made it possible to take non-invasive, fast and detailed tomography scans of the entire colon for the early detection of colon cancer and to conduct routine scans for the early detection of polyps, the precursors to colon cancer.

Acar noted that polyps that form in the colon can turn into malignant tumors within 5-to-15 years and said that it would be possible to prevent colon cancer by detecting and removing them before they turned cancerous.As a result, those at risk for colon cancer should have routine colorectal screening done, he added, and said that this is traditionally accomplished by optical colonoscopy.Acar went on to say, “This is an invasive method that patients usually don’t prefer, which also has medical risks, such as damage to the intestinal mucous membrane.This can be avoided by virtual colonoscopy, where the abdomen CT scan of the patient is used to examine the intestinal surface in detail, a way that is non-invasive.Because examining CT images one-by-one is both costly and open to error, efforts to develop computer technologies that will detect polyps automatically have been ongoing for over two decades in order to increase polyp detection performance and speed up the process.The polyp detection performance of developing technologies is a factor that has a direct impact on the applicability of virtual colonoscopy.These methods have to be both highly sensitive and not miss real polyps, and highly specific and not mark structures that are not polyps as polyps and needlessly direct patients to optical colonoscopy.”

The method developed as part of the “Computer Technologies in Early Colon Cancer Diagnosis and Brain Mapping” research project makes polyps more visible on CT scans and can be effective in boosting the sensitivity and specificity of these technologies.Acar claims that colon cancer can be prevented in a larger number of patients by boosting the success rate of virtual colonoscopy—an early detection method that can be applied during routine controls—should it become a part of clinical practice and says, “This and similar technologies can help to regularly monitor high risk groups and end colon cancer in the future.”