A New Optimization Based Method for Estimation and Spatial Localization of Lung's Air Volume from 3D CT Images

Lung’s air volume estimation is of great importance in lung disease diagnosis. In this paper a fully automatic algorithm, which we presented recently to estimate the lung’s air volume from CT-images, is more developed. In this algorithm, first a suitable cost function is introduced based on the long parenchyma physics to determine the voxels of lung’s air region. In this paper, a fully automatic framework is proposed to calculate the initial guess for the solution of the optimization problem. Moreover, a 3D model reconstruction technique is utilized to determine spatial localization of the lung’s air region in 3D CT-images. Furthermore, the performance of the whole-lung-volume-based methods and direct lung’s air volume measurment methods are compared and investigated. In order to evaluate the accuracy, porcine’s lung images and clinical human’s lung images from reliable databases are fed to the proposed algorithm. The significant accuracy and robust performance of the proposed algorithm is illustrated with respect to the resolution reduction of CT-images.