A perfusion dynamic image segmentation method and system based on a time signal curve

Abstract

The invention discloses a perfusion dynamic image segmentation method based on a time signal curve. The method comprises the following steps: (1) a classifier training step; (1.1) a training data reading step; (1.2) a training data feature extraction step and a training set generation step; (1.3) generating a classifier; (2) using a classifier; (2.1) a step of reading in to-be-analyzed data; (2.2)extracting features of the to-be-analyzed data; And (2.3) classifying and labeling. According to the method, the characteristic parameters of the time signal curves of all voxels in the perfusion data are skillfully and directly utilized, and all the voxels in the perfusion image data are classified and marked, so that classification of different tissues is realized. The invention further discloses a perfusion dynamic image segmentation system based on the time signal curve.

Description

technical field [0001] The invention relates to a perfusion imaging data processing technology, in particular to a segmentation method of a perfusion dynamic image based on a time signal curve. The invention also relates to a segmentation system of the perfusion dynamic image based on the time signal curve. Background technique [0002] With the development of medical imaging technology and computer technology, dynamic brain perfusion imaging and quantitative analysis technology for the purpose of observing the blood perfusion of brain tissue in the medical field have begun to be used in clinical medical auxiliary diagnosis. Through the quantitative analysis of a temporal image sequence obtained by the dynamic cerebral perfusion examination, the blood perfusion of the brain tissue in the region of interest can be measured, and the corresponding time-density curve can be drawn. According to the principle of central volume, these curves are further analyzed and calculated, an...

AI Technical Summary

Problems solved by technology

However, as a result of the reduction of perfusion data, the parametric image substantially loses a lot of useful information, resulting in the inability to obtain stable, effective and accurate results when determining the lesion area only based on the parametric image

In addition, during the calculation of the parameter map, it is necessary to make assumptions about parameters such as the red blood cell content of large and small blood ...

Images