Method for positioning organs in medical image

Abstract

The invention discloses a method for positioning organs in a medical image. The method comprises following steps of step S1, inputting the medical image containing a plurality of slice images; step S2, preprocessing the input images; filtering non-body pixels; step S3, calculating the number of connected domains of each layer of slice image; through judging the numbers and the positions of the connected domains, removing the parts except first organs; confirming that the medical image contains the first organs; calculating the feature value of each layer of slice image, thus obtaining a plurality of feature values; forming a feature curve by the plurality of feature values and the slice number of the slice images; wherein the step of calculating the feature value comprises with respect to each layer of slice image, calculating the proportion of a pixel area of the pixel gray value or CT value relative to the whole body pixel area in a first range; and positioning the slice images in which the organs are located according to the proportions. According to the settings, the ranges in which the organs are located can be positioned rapidly, and the accuracy is high.

Description

technical field [0001] The invention relates to image processing in the medical field, in particular to a method for locating organs and tissues in a three-dimensional CT image. Background technique [0002] Existing methods for identifying and locating organs and tissues in 3D CT images, for example, methods based on machine learning, this method involves complex preliminary work, involves a large number of training image collection and preprocessing, and requires the design of complex classifiers, and the computational complexity of identification and locating higher. In many image processing applications, such as segmentation, registration, automatic recognition of image parts, rough image alignment, etc., it is necessary to perform preliminary identification and judgment on image parts or organs and tissues. Easy way to do it. [0003] Therefore, it is necessary to improve the existing methods for locating organs and tissues in three-dimensional CT images, and to impro...

AI Technical Summary

Problems solved by technology

[0002] Existing methods for identifying and locating organs and tissues in 3D CT images, such as methods based on machine learning, require complex preliminary work, involving a large number of training image collection and preprocessing, and require the design of complex classifiers, and the computational complexity of identification and locating higher

Images