Pulmonary Vascular Segmentation Algorithm Based on Fractional Differential Enhancement

Abstract

With one of the highest incidence rates, Lung cancer accounts for 26.8% of cancer deaths. Due to the pulmonary segment having a separate pulmonary artery and pulmonary vein, the division of the pulmonary artery and pulmonary vein is critical for path planning in surgery of segmental resection of the lung. As a result of this previous research, CT images provided are used as data sources. Combined with the anatomical structure characteristics of pulmonary vessels, the following segmentation methods of the pulmonary artery and pulmonary vein are studied in this thesis. (a) image segmentation and center line extraction is introduced to lung parenchyma segmentation, based on the 3D growth region method. (b) Morphology used to fill the holes in the segmented pulmonary parenchyma, operations made between filled images and the original image to obtain the lungs filled with pulmonary vessels, fractional differential operator designed to enhance the gray level of the small blood vessels, and complete pulmonary blood vessel segmented by local optimal threshold. (c) Algorithm for extracting the centerline of lung blood vessels based on two distance field is researched and arteriovenous conglutination part of the pulmonary centerline detected and removed automatically to obtain the vascular subtree. Finally, Using topological anatomy of the pulmonary artery and vein, the vascular tree classification algorithm based on the subtree leaf node matching is studied along with the pulmonary blood vessels which are grown by region growth method through the center line. The pulmonary artery and vein are then extracted from the chest CT image. The experimental results show that this proposed pulmonary artery and vein segmentation method has high accuracy and short operation time, and has good clinical application value. This newly developed algorithm and implementation will have important applications for treatment planning, dose calculations and treatment validation during cancer radiation treatment.