Segmentation and Reconstruction of the Pulmonary Parenchyma

This work presents a method for the segmentation and reconstruction of the pulmonary parenchyma in CT images. The identification and outline of the pulmonary parenchyma is essential to analyze the pulmonary volume and is an important stage in automatic or semi-automatic processes for detecting lesions, vessels and the bronchial tree. The proposed method here combines traditional concepts, such as global-threshold segmentation, mathematical morphology, border detection and chain code, with new ideas, such as area classification of pulmonary structures and border reconstruction, when the lesion is adjacent to the pleura. We performed tests with 150 computed-tomography images and observed that 135 of them provided good results. Resumo. Neste trabalhoapresentado um m ´ etodo para a segmentac¸ ˜ ao

[1]  T. Michael Cannon,et al.  Image segmentation applied to CT examination of lymphangioleiomyomatosis , 1994, Medical Imaging.

[2]  W J Kostis,et al.  Computer-aided diagnosis for lung cancer. , 2000, Radiologic clinics of North America.

[3]  L.H. De Figueiredo,et al.  IUP/LED: A Portable User Interface Development Tool , 1996 .

[4]  H. K. Huang,et al.  Knowledge-Based Lung Nodule Detection from Helical CT , 1999, Bildverarbeitung für die Medizin.

[5]  Noboru Niki,et al.  Three-dimensional analysis of lung areas using thin slice CT images , 1996, Medical Imaging.

[6]  Y. Kawata,et al.  Computer-aided diagnosis for pulmonary nodules based on helical CT images , 1998, Proceedings. Fourteenth International Conference on Pattern Recognition (Cat. No.98EX170).

[7]  E. Feuer,et al.  SEER Cancer Statistics Review, 1975-2003 , 2006 .

[8]  Anil K. Jain.  Fundamentals of Digital Image Processing , 2018, Control of Color Imaging Systems.

[9]  Noboru Niki,et al.  Three-dimensional analysis of lung areas using thin-slice CT images , 1996, Proceedings of 13th International Conference on Pattern Recognition.