A method and system for detecting pathogens of vaginitis based on attention mechanism and multi-scale fusion

By employing an attention mechanism and multi-scale fusion-based method for detecting vaginal pathogen cells, and utilizing an improved attention mechanism and multi-scale fusion model based on the YOLOv11 model, the accuracy problem of detecting vaginal pathogen cells was solved, enabling rapid and accurate automated diagnosis and pathogen cell count.

CN122116347APending Publication Date: 2026-05-29CHINA UNIV OF MINING & TECH
View PDF 0 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
CHINA UNIV OF MINING & TECH
Filing Date
2026-01-21
Publication Date
2026-05-29

AI Technical Summary

Technical Problem

Current technologies for detecting vaginal pathogen cells have a high risk of false negatives/positives, making it difficult to achieve efficient and accurate automated diagnosis.

Method used

A method for detecting vaginal pathogen cells based on attention mechanism and multi-scale fusion was adopted. By training a vaginal pathogen cell detection model, the attention mechanism and multi-scale fusion model improved by YOLOv11 model were used, and feature extraction and fusion were performed by combining ODDA, CSMSPP and C2PSA modules. The model was optimized by classification, localization and confidence loss to achieve rapid identification of pathogen cells.

Benefits of technology

It improves the detection accuracy of pathogen cells causing vaginitis, reduces the probability of missed and false detections, provides a reliable basis for automated assisted diagnosis, and supports the statistical analysis of pathogen cell counts and diagnosis and treatment.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN122116347A_ABST
    Figure CN122116347A_ABST
Patent Text Reader

Abstract

The application discloses a kind of based on attention mechanism and multi-scale fusion's vaginitis pathogen cell detection method and system.The method will be obtained in advance vaginal secretion immunofluorescence microscopic image input training completed optimal pathogen detection model, output the detection result of recognition trichomonad, leukocyte, epithelial cell, clue cell and spore.The application is based on deep learning algorithm to realize the automatic high-efficiency accurate detection of vaginitis pathogen cell, improve the confirmed diagnosis rate of vaginitis, greatly reduce the probability of missed detection, misdiagnosis phenomenon in diagnosis process.According to pathogen cell detection result, simulate clinical diagnosis output, realize pathogen cell classification judgment, provide reliable basis for the automation auxiliary diagnosis of vaginitis.
Need to check novelty before this filing date? Find Prior Art