Method for fast clustering medical sequential images

Abstract

A method for fast clustering a medical sequential images belongs to the technical field of computer aided diagnosis of medical imaging, and comprises the following steps of: preprocessing, forming a feature vector matrix, forming a target vector matrix, computing a linear transformation matrix, measuring a space vector feature space transformation, clustering a feature space, and clustering a measuring space. The method has high image processing speed, high clustering precision and strong clinical utility.

Description

technical field [0001] The invention relates to the technical fields of medical image processing and imaging computer-aided diagnosis, in particular to a fast clustering method of medical sequence images. Background technique [0002] With the improvement of clinical application requirements of imaging diagnosis and the development of medical imaging technology, medical imaging data more generally exists in the clinic in the form of massive, two-dimensional or three-dimensional sequence images. For medical sequence images, doctors are often only interested in one or several types of substances, which makes it particularly clinically meaningful to seek related clustering methods. According to the investigation, so far, a small number of researchers have studied the clustering of medical sequence images. For example, Jiawan Zhang et al. in the article "Automatic Classification of MRI Images for Three-dimensional Volume Reconstruction by Using General Regression Neural Networks...

AI Technical Summary

Problems solved by technology

[0003] Aiming at the disadvantages of the previous methods described in the background technology, such as slow processing speed, low clustering accuracy, and poor clinical practicability, the present invention proposes a fast clustering method for medical sequence images, which first linearly transforms the original sequence image data in the measurement space into In the feature space, clustering is performed in the feature space, and finally the clustered data in the feature space are mapped back to the measurement space one by one to realize the clustering of the measurement space

Images