关键词: 全自动, 三维分割, 肺实质

Abstract: The precise three-dimensional（3-D） segmentation of pulmonary parenchyma from thorax X-ray computed tomography（CT） images is essential for the detection of pulmonary diseases.The complex 3-D structure of breast tissues in CT images makes precise segmentation difficult.A fully automatic 3-D segmentation algorithm is presented，which based on threshold，region growing and improved wavefront simulation model for extracting the 3-D pulmonary parenchyma from a X-ray CT dataset.The new algorithm is applied on the Computer Vision and Image Analysis Group/International Early Lung Cancer Action Program（VIA/I-ELCA） and Lung Image Database Consortium（LIDC） database.The results indicate that the algorithm is fast and precise for pulmonary parenchyma segmentation.

Key words: fully automatic, 3D segmentation, pulmonary parenchyma