医学数据智能识别方法和装置
By optimizing the feature robustness and noise resistance of the intelligent recognition model, the problems of data imbalance and annotation noise are solved, and the recognition accuracy and stability of the model in fields such as medical image analysis are improved.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- BEIJING XIAOYING TECH CO LTD
- Filing Date
- 2025-10-24
- Publication Date
- 2026-07-17
AI Technical Summary
Existing intelligent recognition methods have failed to effectively address issues such as data imbalance, annotation noise, and cross-domain distribution, resulting in poor recognition performance of models in fields such as medical image analysis.
By constructing optimization objectives of minimizing empirical risk, feature robustness, and label noise denoising, the intelligent recognition model is optimized, and an adaptive noise-resistant loss function is introduced to improve the model's feature robustness and noise resistance.
It improves the model's recognition accuracy and stability under imbalanced data and cross-domain distribution, enhances the model's generalization ability and noise resistance, and significantly improves recognition performance in rare categories and complex scenes, especially in medical image analysis.
Smart Images
Figure CN121682476B_ABST