A system and method for multiple classification of breast ultrasound images based on cross-correlation features

Abstract

The invention provides a system and method for multiple classification of breast ultrasound images based on cross-correlation features, including an ultrasound image preprocessing unit; a region of interest extraction unit for extracting images of the region of interest; an internal cross-correlation density feature extraction unit for extracting sense The internal cross-correlation density eigenvalue of the image of interest; the traditional feature extraction unit is used to extract a variety of traditional eigenvalues ​​of the image of the region of interest; the multi-class classification unit is used to train the classifier and convert the internal cross-correlation density eigenvalue vector The traditional feature vectors are input to the three trained classifiers for classification, and the most predicted category is taken as the final classification result. The classification method of the present invention increases the feature based on the internal cross-correlation density of the region of interest, can effectively improve the effect of breast ultrasound computer-aided diagnosis, and increases the classification category of breast cysts, a benign lesion, and further satisfies doctors' requirements for breast ultrasound computer-aided systems. demand.

Description

technical field [0001] The invention belongs to the technical field of post-processing of medical images, and in particular relates to a system and method for multiple classification of breast ultrasound images based on cross-correlation features. Background technique [0002] Breast disease is one of the more common diseases in modern women, and breast cancer is a common female malignant tumor. In terms of incidence, breast cancer ranks first among female malignant tumors in Europe and the United States, and ranks second in China. It has become an important disease that affects women's physical and mental health. Regarding the differentiation of benign and malignant breast masses, imaging methods undoubtedly play an important role. Imaging examinations such as mammography, CT, MRI, and ultrasound all have an important impact on the diagnosis and differentiation of breast masses. Among them, with the update of equipment and the development of technology, more and more atten...

AI Technical Summary

Problems solved by technology

[0003] However, medical ultrasound images are coherent images. Compared with mammography X-ray images, CT, and MRI, there are still defects such as low image quality, extremely obvious artifacts, and low signal-to-noise ratio that easily affect the diagnostic results.

Images