Method for extracting terminal bronchial tree from lung CT image

Abstract

A method for extracting a terminal bronchial tree from a lung CT image belongs to the field of image processing techniques based on medical science images. The method includes the steps of removing lung area edges in a lung CT image through an area growth method, searching a suspected bronchial tree area according to CT value in a lung area with removed edges, removing non-tracheae tree area in the suspected bronchial tree area to obtain a terminal bronchia, obtaining two end points of the terminal bronchia, generating a center line of the terminal bronchia according to the two end points of the terminal bronchia, and connecting the terminal bronchia with a tracheae tree body along the center line of the terminal bronchia to obtain a bronchial tree. Terminal bronchi are easily missed and not segmented in three-dimensional bronchial tree segmentation, and missing tracheae points are extracted according to CT values and screened according to the morphological characteristics of a tracheae tree and connected to the tracheae tree along the missing tracheae extension lines. The terminal bronchi can be segmented more accurately, and a more complete three-dimensional bronchial tree can be obtained.

Description

technical field [0001] The invention belongs to the technical field of image processing based on medical images, and in particular relates to a method for extracting terminal bronchus trees from lung CT images. Background technique [0002] Lung tomographic images obtained by computed tomography (CT) scanners contain a large amount of information that can reflect the intrinsic pathophysiology of lung diseases. Some of these messages manifest as structural abnormalities of the bronchial tree in the lungs, such as reduced bronchial lumen area and thinner wall thickness. The visual inspection commonly used in clinical practice cannot provide a quantitative description of the airway tree, and manual extraction of the airway tree is a time-consuming and huge work. Using computer image processing and analysis technology, automatic segmentation and extraction of three-dimensional bronchial trees from CT images is the basis for subsequent quantitative analysis, which is expected to...

AI Technical Summary

Problems solved by technology

However, this assumption also leads to two general problems of airway tree segmentation: leaky and leaky segmentation

The Chinese patent application number CN104504737A "A Method for Obtaining a Three-dimensional Tracheal Tree from CT Images of the Lungs" proceeds from a new perspective, first segmenting the main bronchi and bronchial segments, and then dividing each segment of the bronchi and the main bronchi stitching", but there are a large number of fine terminal bronchi that have not been segmented, and a large number of complex "stitching" algorithms will inevitably lead to increased computational costs

[0005] In summary, the segmentation method of the main bronchial tree has been perfected at present, but the extraction and connection of the terminal bronchus is still a big problem. The present invention aims to better solve this problem

Images