Lung lobe segmentation method and system based on space, storage medium and electronic equipment

Abstract

The invention discloses a lung lobe segmentation method and system based on space, a storage medium and electronic equipment, and relates to the field of lung lobe segmentation. The system comprises an input module, a segmentation module, a space circulation module and an output module, the input module is used for obtaining a CT scanning image of a lung and inputting the CT scanning image to thesegmentation module, the segmentation module is used for processing the received CT scanning image and inputting a processing result to the space circulation module, The space circulation module is used for carrying out superposition circulation processing on the received processing result and outputting the superposition circulation processing result to the output module, and the output module isused for outputting the received superposition circulation processing result. According to the method, the problem that the pulmonary lobe is accurately segmented without deducing the interlobe fracture position according to the front layer and the rear layer of the CT image can be solved, and the effects of gentle pulmonary lobe transition and reasonable pulmonary lobe range prediction can be achieved.

Description

technical field [0001] The invention relates to the field of lung lobe segmentation, in particular to a space-based lung lobe segmentation method, system, storage medium and electronic equipment. Background technique [0002] The traditional interactive method requires manual participation and cannot achieve automatic lung lobe segmentation; the pure deep learning method only considers the spatial correlation, but the lung lobe segmentation task is a time-sequence-related task, and the deep learning segmentation based on convolutional neural network methods, usually only focus on two-dimensional spatial information, although in medical image processing, often input multiple adjacent sequences at a time or use three-dimensional input, but this method only adds time series information to the input data, in most interlobar fissures When it is not clear, the position of the interlobar fissure is not inferred from the front and back layers of the CT image, and the lung lobes are ...

AI Technical Summary

Problems solved by technology

[0002] The traditional interactive method requires manual participation and cannot achieve automatic lung lobe segmentation; the pure deep learning method only considers the spatial correlation, but the lung lobe segmentation task is a time-sequence-related task, and the deep learning segmentation based on convolutional neural network methods, usually only focus on two-dimensional spatial information, although in medical image processing, often input multiple adjacent sequences at a time or use three-dimensional input, but this method only adds time series information to the input data, in most interlobar fissures In unclear cases, the position of the interlobar fissure was not inferred based on the front and back layers of the CT image, and the lung lobes were segmented accurately

Images