用于肺炎影像判读的融合辅助决策模型构建方法及系统
By constructing an inner-layer iterative update of the health code book and counterfactual images, and combining the consistency loss function to optimize the parameters of the interpreter and locator, the problem of loose correspondence between interpretation conclusions and lesion hot zones in existing pneumonia image interpretation models is solved, and stable and consistent output under various conditions is achieved.
CN122023405BActive Publication Date: 2026-07-17ANHUI UNIV
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- ANHUI UNIV
- Filing Date
- 2026-04-13
- Publication Date
- 2026-07-17
Smart Images
Figure CN122023405B_ABST
Abstract
本发明公开了用于肺炎影像判读的融合辅助决策模型构建方法及系统，涉及智能辅助诊断技术领域，现提出如下方案，其包括获取肺炎样本影像与健康肺部影像，并构建判读器与定位器，对健康肺部影像提取肺实质局部纹理并聚类离散化，得到表征健康纹理的离散向量集作为健康码本，利用定位器识别肺炎样本影像中的病灶区域。本申请通过构建健康码本并生成反事实健康影像，再以预设健康阈值触发的内层迭代将当前判读值反馈到定位器参数更新中，解决了现有方案中判读结论与病灶热区松散对应且易受与病灶无关背景线索干扰而导致可信度不足的缺陷，实现了判读结论可由反事实对照进行校验且定位证据与判读依据一致的可解释输出。
Need to check novelty before this filing date? Find Prior Art