Pulmonary emphysema image processing method and system based on low data requirements

Abstract

The invention provides a pulmonary emphysema image processing method and system based on low data demands. The method comprises the following steps: M1, preparing a pulmonary CT film marked with the yin and yang of a pulmonary emphysema focus, and forming a group of medical digital imaging and communication files; m2, preprocessing the prepared lung CT film, and obtaining a three-dimensional arraythrough a group of medical digital imaging and communication files; m3, constructing a deep convolutional neural network architecture, training a deep convolutional neural network through the three-dimensional data, and judging a pulmonary emphysema image through the deep convolutional neural network; required characteristics can be automatically learned from chest CT with emphysema yin-yang marks, and image processing yin-yang judgment is carried out. Compared with a common CT deep neural network image processing auxiliary diagnosis technology, the technology avoids the problems that a 3D model occupies a large amount of memory and is poor in performance on a CT with a thick layer, also avoids the limitation that a 2D model cannot comprehensively utilize three-dimensional space information, and fully utilizes the spatial relationship between layers.

Description

technical field [0001] The present invention relates to the field of medical imaging, in particular, to a method, system and medium for emphysema image processing based on low data requirements, especially to a high-speed and lightweight chest CT image emphysema image processing method and system based on low data requirements And the medium, the way of automatic processing of chest CT images and its training method based on deep convolutional neural network trained with weakly labeled medical image data. Background technique [0002] CT images belong to the category of 3D medical imaging. For the current computer-aided diagnosis model based on deep neural networks, the biggest challenge lies in the image data represented by each pixel in the three dimensions of length, width and layer depth. The physical spatial distance (spacing) is not uniform (anisotropic). Usually, each pixel in the layer dimension represents several times the distance in the length and width dimensions...

AI Technical Summary

Problems solved by technology

This makes the neural network architecture more complex, the training more time-consuming and laborious, the memory usage increases significantly, and the model reasoning speed drops significantly

Images