一种基于口腔影像数据的牙齿状态评估方法、装置及系统

By acquiring multi-angle oral cavity images and peripheral data during tooth detection, and using backbone networks and path aggregation networks for feature extraction and fusion, the problem of insufficient precision and weak domain knowledge integration in existing technologies is solved, thus achieving efficient tooth condition assessment.

CN122415494APending Publication Date: 2026-07-17SHANXI MEDICAL UNIV
View PDF 0 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
SHANXI MEDICAL UNIV
Filing Date
2026-04-20
Publication Date
2026-07-17

AI Technical Summary

Technical Problem

Existing dental detection technologies suffer from insufficient precision, weak integration of domain knowledge, and poor coupling between multiple tasks in multi-angle and multi-surface detection scenarios, making it difficult to achieve fine segmentation at the tooth surface level and effectively utilize prior knowledge in oral medicine.

Method used

By acquiring multi-angle oral cavity images and peripheral data, a backbone network is used to extract multi-scale feature maps, and feature calibration is performed by combining channel attention weight vectors. A path aggregation network is used for feature fusion, and a composite loss function is used to optimize the model, thereby achieving efficient detection for multiple tasks.

Benefits of technology

It improves the accuracy and efficiency of tooth condition assessment, enabling comprehensive detection of macroscopic defects ranging from minute caries to dental arch morphology, and reducing missed detections due to lack of perspective.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN122415494A_ABST
    Figure CN122415494A_ABST
Patent Text Reader

Abstract

本发明涉及牙齿检测技术领域,尤其涉及一种基于口腔影像数据的牙齿状态评估方法、装置及系统;通过将多角度口腔图像与至少包括年龄的外围数据结合,使模型在特征提取阶段即融入患者个体信息,以提升牙齿状态评估的准确性,通过骨干网络中级联的C2f模块和SPPF模块生成至少三个尺度的特征图,并将不同分辨率特征图分别与不同的检测任务进行关联,同时检测头采用双向特征融合路径实现跨尺度的信息互补,实现了从微小龋坏等细微缺陷到牙弓形态等宏观缺陷的全面检测,通过引入分布焦点损失,与CIoU损失、二元交叉熵损失共同构成复合损失函数,将边界框坐标回归建模为离散概率分布,结合多角度口腔图像有效降低了因视角缺失导致的漏检。
Need to check novelty before this filing date? Find Prior Art