Method for positioning and segmenting cerebellar earthworms in ultrasonic image

Abstract

The invention discloses a method for positioning and segmenting a cerebellar earthworm part in an ultrasonic image. The comprises the following steps of: obtaining a human brain ultrasonic image and corpus callosum size and position information in the ultrasonic image, reading a cerebellar earthworm part image, and generating a cerebellar earthworm part average template image through an adversarial neural network; according to the corpus callosum position information, an area with a small pixel mean value is selected as an initial search area of the cerebellar earthworm part, reducing the position information Loc of the cerebellar earthworm part relative to the corpus callosum direction; taking the average template size of the adaptive cerebellar earthworm as the size of the initial searchresult of the cerebellar earthworm; performing adaptive template size sliding window search by adopting a new similarity comparison algorithm so as to obtain a cerebellar earthworm fine search result; and carrying out outer contour graph fitting on the direction position information Loc and the morphological characteristics of the cerebellar earthworm part to obtain an initial contour of the cerebellar earthworm part, and carrying out smoothing processing on the accurate contour to finally obtain cerebellar earthworm part contour information.

Description

technical field [0001] The invention relates to the field of medical image processing, in particular to a method for positioning and segmenting cerebellar vermis in an ultrasonic image. Background technique [0002] With the wide application of computer technology in medicine, the earlier manual segmentation of images is gradually replaced by computer automatic segmentation. At present, how to detect whether the fetal cerebellar vermis is developing normally, usually manually locate and segment the cerebellar vermis area in the ultrasound image, and calculate its outline length, area and other characteristic information to detect whether the development is normal. How to accurately and efficiently locate the vermis of the cerebellum, automatically segment it, and extract position and contour information has become a key issue. At present, there is no known effective computer automatic processing method for the problem of ultrasonic image positioning and segmentation of the ...

AI Technical Summary

Problems solved by technology

At present, there is no known effective automatic computer processing method for the problem of ultrasonic image positioning and segmentation of cerebellar vermis, and it is all manually positioned and segmented

This requires professional knowledge reserves, and it takes a lot of time, and the effect is not necessarily fine, so the accuracy cannot be guaranteed

Images