Automated Detection System Of Peripheral Pulmonary Carcinoma In CT Images

Medicine-image-based Computer-aided Diagnosis, that is which aims at individual diversity of diseases, deals with very complex problems. Basically, it sums up the foundations of medicine image processing and pathological characteristics of neoplasm.When interpreting, the radiotherapist makes qualitative analysis of the focus based on the signs of it in CT films. Generally speaking, if a focus has the signs, that is lobulation sign, cavitation sign, speculation sign, pleural indentation, it will be determined as peripheral carcinoma of the lung. Much experience is a must for making accurate analysis of lobulation sign andcavitation sign. And, withal, the analyst brings many subjective factors. The system of computer-aided diagnosis is able to provide objective conclusion for doctors. As a system of computer-aided diagnosis, the automated detection system of peripheral pulmonary carcinoma based on CT images aims to supply a reference for making radiation treatment. A series of CT images is determined as the input of the system, and then the system output the contour of a region what is determined as peripheral pulmonary carcinoma. Certainly, we must preprocess the images, extract feature of the segmented region and estimate the quality of the region.The part of preprocessing involves neighborhood average, image segmentation and aided operate.With image segmentation, we extract the contour of the skin, the lung and the region of interest. We get the contour of skin by segmenting the image with an...