A method for classifying benign and malignant breast lesions

By constructing a multimodal image feature fusion model and a text feature extraction model for classifying benign and malignant breast lesions, the problems of difficulty in multimodal image feature fusion and insufficient classification accuracy in radiological images are solved, and a higher accuracy classification of benign and malignant breast lesions is achieved.

CN120747639BActive Publication Date: 2025-10-31THE SECOND XIANGYA HOSPITAL OF CENT SOUTH UNIV
2 Cites 0 Cited by

Patent Information

Application Number
CN202511187856.9
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2025-08-25
Publication Date
2025-10-31
Estimated Expiration
2045-08-25

AI Technical Summary

Technical Problem

Existing technologies for classifying benign and malignant breast lesions based on radiographic images suffer from difficulties in fusing multimodal image features and low classification sensitivity, especially due to insufficient classification accuracy caused by the differences between image features and text features and the lack of microstructural features.

Method used

A classification model for benign and malignant breast lesions was constructed, including a multimodal image feature fusion model, a text feature extraction model, a common semantic space, and a mask prediction model based on a stacked multi-layer bidirectional Transformer decoder. By using contrastive learning and supervised training methods, image and text features were fused to improve classification accuracy.

Benefits of technology

It effectively improves the segmentation accuracy of breast lesions in radiological images and the accuracy of benign and malignant classification, providing more reliable classification results and has potential clinical application value.

✦ Generated by Eureka AI based on patent content.
Patent Text Reader

Abstract

This invention provides a method for classifying benign and malignant breast lesions, comprising the following steps: Step 1, obtaining N sample pairs; Step 2, constructing a benign and malignant breast lesion classification model; Step 3, training the benign and malignant breast lesion classification model; Step 4, obtaining the benign and malignant breast lesion classification results. This invention can effectively improve the segmentation accuracy of breast lesions in radiological images, providing more accurate and reliable benign and malignant classification results, and has potential clinical application value.
Need to check novelty before this filing date? Find Prior Art