Lung bronchial segmentation and calibration method

Abstract

The invention discloses a lung bronchial segmentation and calibration method. The method comprises the following steps of S1, extracting a lung airway tree region from image data; S2, setting an interested voxel region according to the lung airway tree region, and performing sharpening filtering processing on the interested voxel region; S3, classifying the voxel regions of interest through a deeplearning convolutional neural network according to the processed voxel regions of interest, removing leakage and tracking airway branches to obtain an optimized lung airway tree region; S4, extracting airway tree structure information from the optimized lung airway tree region, and obtaining a candidate name set of each airway and a weight of each candidate name through a lung airway name labeling model according to the lung airway tree region and the airway tree structure information; and S5, optimizing the names of each group of airway trees according to the weights, and implementing accurate and automatic labeling of the names of the airway trees to obtain airway name information.

Description

technical field [0001] The present invention relates to a segmentation and calibration method of pulmonary bronchi, in particular to a precise and automatic segmentation of pulmonary airways based on voxel sharpening tracking and deep learning convolutional neural network and a lung airway segmentation based on deep learning convolutional neural network. A method for automatic labeling of bronchus names. Background technique [0002] The existing lung airway segmentation method is mainly based on the adaptive region growth method, that is, the automatic or manual selection of the seed point, starting from the seed point, selecting candidate points in the 6 neighborhood or 26 neighborhood, according to the growth criterion Determine whether the candidate point is the target point, and then use the target point as a new seed point for the next adaptive growth. This type of method is implemented by a variety of different algorithms, and the difference between these algorithms ...

AI Technical Summary

Problems solved by technology

This increases the difficulty of interventional surgery, surgical risk and operation time, and also leads to surgical misoperation

Images