Semi-supervised medical image segmentation method based on high-value region adaptive learning

By constructing a semi-supervised medical image segmentation model with adaptive learning of high-value regions, and utilizing confidence-guided cross-prototype consistency learning and dynamic teacher competition teaching modules, the problem of ignoring regional value differences in existing methods is solved, and the accuracy of medical image segmentation and the adaptability to complex regions are improved.

CN120807545APending Publication Date: 2025-10-17SHAANXI UNIV OF SCI & TECH
0 Cites 1 Cited by

Patent Information

Application Number
CN202510816340.X
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2025-06-18
Publication Date
2025-10-17

AI Technical Summary

Technical Problem

Existing semi-supervised learning methods fail to effectively identify and learn the value differences between different regions in medical image segmentation, ignore the learning strategies of key areas, and thus limit the segmentation performance of the model in complex anatomical structures and pathological changes.

Method used

A semi-supervised medical image segmentation model based on adaptive learning of high-value regions is constructed. Two parallel networks are used to impose random perturbations in the feature space to screen out reliable unstable and unreliable stable regions. Differentiated learning strategies are designed, including a confidence-guided cross-prototype consistency learning module and a dynamic teacher competition teaching module, to improve the segmentation performance of the model in key areas.

Benefits of technology

It effectively improves the model's recognition ability in key areas, reduces the negative impact of noise and uncertainty, enhances the model's adaptability to complex areas, and significantly improves the accuracy of medical image segmentation.

✦ Generated by Eureka AI based on patent content.
Patent Text Reader

Abstract

The invention discloses a semi-supervised medical image segmentation method based on high-value region adaptive learning, and the method comprises the steps: constructing a semi-supervised medical image segmentation model based on high-value region adaptive learning, and the model comprises two parallel networks which have the same structure and independently update parameters; the two parallel networks output an original prediction result and a disturbed prediction result, and different value regions, namely a reliable and unstable region and an unreliable and unstable region, are screened out based on confidence distribution of the disturbed prediction result; training a reliable and unstable area through a confidence-guided cross prototype consistency learning module, and training an unreliable and unstable area through a dynamic teacher competition teaching module; and training the model by using the training set, selecting an optimal weight based on the verification set, inputting the test set into the model with the optimal weight, and outputting a result. Through the differential learning strategy, the method is accurately focused on the high-value region of the medical image, and the segmentation effect of the model in the key region of the image is remarkably improved.
Need to check novelty before this filing date? Find Prior Art