一种卵丘细胞剥除质量评估方法及装置
By acquiring images of cumulus cells at different focusing levels using deep learning technology, and using feature extraction and fusion structures for cumulus cell identification and pairing, the subjective error and instability problems of cumulus cell removal quality assessment in existing technologies are solved, and accurate quantification and automatic assessment of cumulus cell removal quality are achieved.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- SUZHOU BOUNDLESS MEDICAL TECH CO LTD
- Filing Date
- 2025-08-06
- Publication Date
- 2026-07-17
AI Technical Summary
Existing methods for assessing the quality of cumulus cell removal rely on manual observation or image thresholding, which suffer from large subjective errors or unstable segmentation and recognition, and cannot accurately quantify the quality of cumulus cell removal.
Deep learning technology is used to acquire cumulus cell images at different focusing levels through YOLO and Siamese or SwAV models. Res2Net-COT residual structure and bidirectional feature pyramid structure are used for feature extraction and fusion. Cell identification and pairing are performed by combining spatial pyramid pooling layer and feature band structure to achieve quantitative evaluation of cumulus cell removal quality.
It enables objective and accurate assessment of the quality of cumulus cell removal, reduces the impact of changes in microscope field of view depth and image quality, improves the automation and accuracy of assessment, enhances the ability to identify cumulus cell features, and reduces noise interference.
Smart Images
Figure CN120976153B_ABST