A STUDY OF LOCAL SLICE FEATURES FOR A LUNG OPACITIES CLASSIFICATION IN THIN-SECTION COMPUTED TOMOGRAPHY IMAGES

摘要

A lung opacities classification in thin-section computed tomography(CT) images is expected for designing a computeraided diagnosis(CAD) system. In classifying lungopacities, the effectiveness of slice features has beenshown. In this paper, in order further to improve the lungopacities classification performance, we have presented alocal slice feature based method.