Computer aided diagnosis system for lung cancer based on helical CT images

In this paper we describe a computer assisted automatic diagnosis system for lung cancer that detects tumor candidates at an early stage from helical computerised tomographic (CT) images. This automation of the process reduces the time complexity and increases the diagnosis confidence. Our algorithm consists of an analysis part and a diagnosis part. In the analysis part, we extract the lung and pulmonary blood vessel regions and analyze the features of these regions using image processing techniques. In the diagnosis part, we define diagnosis rules based on these features, and detect tumor candidates using these rules. We have applied our algorithm to 450 patient's data for mass screening. The results show that our algorithm detected lung cancer candidates successfully.