Lung airway tree segmentation method for tracheal wall notch plugging under sphere-like area constraint

Abstract

The invention discloses a lung airway tree segmentation method for tracheal wall notch plugging under sphere-like area constraint, and belongs to the field of image processing. In order to solve the problem of leakage caused by a low-brightness trachea wall notch in lung airway tree segmentation, the invention provides the lung airway tree segmentation method for plugging the trachea wall notch under sphere-like area constraint. The method comprises the following steps: (1) establishing a sphere-like area constraint model; (2) carrying out intra-area segmentation and parameter self-adaptive adjustment; and (3) plugging a trachea wall notch. The difficulty of leakage detection is reduced due to the constraint of the sphere-like area on segmentation, all parameters can be adjusted easily ina self-adaptive mode in the constraint area, and conditions are created for trachea wall notch plugging through the spheroid area constraint; tracheal wall notch plugging can effectively position andplug a tiny low-brightness tracheal wall notch, and a good segmentation environment is created for extraction of the three-dimensional structure of the pulmonary airway tree.

Description

technical field [0001] The invention belongs to the field of medical image processing, in particular to lung airway tree segmentation based on chest CT images. Background technique [0002] An accurate three-dimensional pulmonary airway tree model constructed based on chest 3D-CT image data is of great significance for the auxiliary diagnosis of lung diseases. In addition to the common problems in medical image segmentation, the research on lung airway tree segmentation algorithm also needs to face two unique problems, namely, the leakage problem and the small airway extraction problem. The root of the two problems is that under the influence of partial volume effect and noise, there are low-brightness tracheal wall gaps on the high-brightness tracheal wall in CT images, such as figure 1 As shown, it is difficult for the tracheal wall to perfectly isolate the airway and lung parenchyma, resulting in the segmentation process being mistakenly extended from the airway to the l...

AI Technical Summary

Problems solved by technology

[0003] In view of the unique leakage problem and small airway extraction problem in the field of pulmonary airway tree segmentation, the main reason for the difficulty of pulmonary airway tree segmentation is considered that the gap in the low-brightness tracheal wall is considered

Images