A cross-modality medical image segmentation method based on pathological anchoring

CN122289680APending Publication Date: 2026-06-26LANZHOU JIAOTONG UNIV
View PDF 0 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
LANZHOU JIAOTONG UNIV
Filing Date
2026-03-20
Publication Date
2026-06-26

AI Technical Summary

Technical Problem

Existing cross-modal medical image segmentation methods struggle to simultaneously maintain consistency in appearance, texture, and structure between ultrasound and dermoscopy, leading to semantic inconsistencies, boundary shifts, and loss of small target structures in the segmentation results. Furthermore, they are difficult to generate high-quality pseudo-labels when the target domain is unlabeled.

Method used

A pathological anchor-based cross-modal medical image segmentation method is adopted. A bidirectional cross-scale co-encoder is used to achieve bidirectional interaction between low-level details and high-level semantics. Semantic anchors are generated by combining pathological priors, and pseudo-labels are stably generated through uncertainty closure and anchor refinement methods, forming a closed-loop path from structure to semantics to noise control.

Benefits of technology

Under unlabeled constraints in the target domain, medical image segmentation with more accurate boundaries, better calibration, and more robust generalization was achieved, improving the segmentation accuracy and generalization ability of lesion regions, explicitly mitigating semantic drift, and enhancing structural reliability.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN122289680A_ABST
    Figure CN122289680A_ABST
Patent Text Reader

Abstract

This invention discloses a cross-modal medical image segmentation method based on pathological anchoring. Using pathological priors as semantic anchors, this method integrates cross-scale collaboration and uncertainty perception in a closed-loop prompting refinement process. This achieves interpretable suppression and robust learning of cross-modal domain shifts, forming a closed-loop path from structure to semantics to noise control. Under strong cross-modal zero-sample settings such as multi-source ultrasound cross-domain segmentation and color dermoscopy, stable and consistent improvements are achieved, manifested in more precise boundaries, better calibration, and more robust generalization. This provides technical support for disease image diagnosis.
Need to check novelty before this filing date? Find Prior Art