A method, device and equipment for preoperative risk stratification of endometrial cancer

By segmenting, labeling, and encoding preoperative ultrasound images and pathological biopsy data for endometrial cancer, and using a cross-modal attention gating module to generate dynamic weights, the problem of inconsistency between biopsy and ultrasound image assessments was solved, improving the accuracy of preoperative risk stratification and the rationality of surgical plans.

CN122136003APending Publication Date: 2026-06-02XIAMEN XINGLIN HOSPITAL (XIAMEN INFECTIOUS DISEASE HOSPITAL) +1
View PDF 0 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
XIAMEN XINGLIN HOSPITAL (XIAMEN INFECTIOUS DISEASE HOSPITAL)
Filing Date
2026-05-06
Publication Date
2026-06-02

AI Technical Summary

Technical Problem

In existing technologies, when biopsy pathology and ultrasound imaging are evaluated independently in preoperative risk stratification assessment of endometrial cancer, the inconsistencies in the conclusions cannot be automatically identified and the confidence weights cannot be dynamically adjusted, leading to inaccurate risk stratification.

Method used

By acquiring ultrasound imaging data and pathological biopsy data, segmentation, labeling, and structured coding are performed to extract radiomics and pathological feature vectors. A cross-modal attention gating module is used to generate dynamic weight vectors, which are then spliced ​​after feature weighted modulation and input into a multi-task prediction network to output risk stratification results.

Benefits of technology

It enables automatic adjustment of confidence levels when there is discrepancy between biopsy and ultrasound imaging assessments, improving the accuracy of preoperative risk stratification, generating more accurate muscle layer invasion depth and lymph node metastasis risk probability, and recommending reasonable surgical plans.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN122136003A_ABST
    Figure CN122136003A_ABST
Patent Text Reader

Abstract

This invention provides a method, device, and equipment for preoperative risk stratification assessment of endometrial cancer. It involves segmenting and annotating preoperative ultrasound images to extract radiomics feature vectors, and simultaneously performing structured encoding and mapping of pathological biopsy data to generate pathological feature vectors. These two feature vectors are then input into a fusion network containing a cross-modal attention gating module. This gating module automatically generates dynamic weight vectors based on the consistency between the image and pathological features. The two feature vectors are then weighted and modulated separately before being concatenated to obtain a joint feature representation. This automatically reduces the weight of the lower-confidence modality and amplifies the weight of the higher-confidence modality when their assessment conclusions are inconsistent. Finally, the joint feature representation is input into a multi-task prediction network to output the probability of myometrial invasion depth and the probability of lymph node metastasis risk, thereby generating preoperative risk stratification results and surgical plan recommendations.
Need to check novelty before this filing date? Find Prior Art