Implementing a System for Diagnosing Lung Cancer based on Active Contour Algorithms

Lung cancer is a very common cancer, but it only becomes noticeable relatively late in the form of symptoms and complaints. To date, the possibilities of early detection are limited - the examinations tested so far are stressful and not without risk. In particular, people who are at high risk of developing lung cancer should be aware of important warning signs. Due to late detection, often in therapeutically overdosed stages, only 14% of the diagnosed patients are able to survive for a longer period of time (5 years). Once lung cancer begins to cause symptoms, it is usually visible on an X-ray and next, CT scan of the chest may be ordered for a more detailed exam. In this paper we aimed by image processing to develop a system for lung cancer recognition and detect the stage of the disease. In this study, some ultrasonography images with different types of lung cancer and with no lung cancer were tested. Active contour algorithms were designed and applied to the CT images in order to segment the image without clearly defined boundaries. Next, the resulted images have been interpreted. The proposed method for diagnosing lung cancer is useful, because it helps doctors to diagnose if it is the case of lung cancer or other disease.