Colon polyp real-time segmentation method and system fusing multi-scale high-frequency information
By using a real-time segmentation method for colon polyps that integrates multi-scale high-frequency information and enhances the boundary features of colon polyps by utilizing a high-frequency boundary-guided fusion module group, the problem of insufficient segmentation accuracy for small target polyps is solved, achieving efficient and accurate segmentation of colon polyps and reducing false positive and false negative rates.
Patent Information
- Application Number
- CN202511015873.4
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2025-07-23
- Publication Date
- 2025-10-31
- Estimated Expiration
- 2045-07-23
AI Technical Summary
Current colonoscopy examinations have a high rate of missed polyp diagnosis. Existing segmentation techniques are not accurate enough for small polyps and have high false positive and false negative rates, making them difficult to apply effectively in the field of computer vision.
A real-time segmentation method for colon polyps is adopted, which integrates multi-scale high-frequency information. The high-frequency boundary guides the fusion module group, which combines the feature extraction module and the high-frequency boundary mining module. It utilizes edge gradient amplitude and dynamic modulation mechanism to enhance the boundary features of colon polyps and reduce false positive and false negative predictions.
It improves the segmentation accuracy of small-target colon polyps, reduces the probability of false positives and false negatives, and enhances the segmentation effect, making it suitable for the computing efficiency of edge computing devices and servers.