Lung Segmentation in Pulmonary CT Images using Wavelet Transform

Computer-aided diagnosis (CAD) has become a major research interest in diagnostic radiology and medical imaging. The basic goal of CAD is to provide a computer output as a second opinion to assist medical image interpretation by improving accuracy, consistency of diagnosis, and image interpretation time. Since a CAD system is only interested in analyzing a specific organ, segmentation of computer tomography (CT) images is a precursor to most image analysis applications. A fully automated method is presented to segment lung in pulmonary CT images based on detected lung edges by wavelet analysis. Due to wavelet transformation characteristics, the proposed method is not only computational inexpensive compared to other existing methods such as snakes or watershed, but also is robust and accurate in detecting lung borders. A set of 330 low dose (50 mA) CT images were processed demonstrating accuracy and satisfactory performance of the algorithm.