人工智能驱动的乳房造影质量保证:在韩国通过自动化幻影图像评估提高质量控制效率
Hoo Yun1, Sanghyun Noh2, Hyungwook Cho3
1Korea University College of Medicine, Seoul, Republic of Korea.
PloS one
|September 8, 2025
概括
一个深度学习模型准确地评估了乳房镜幻影图像,提高了韩国的质量控制一致性和效率. 这种人工智能工具在评估人工损伤以获得认证时实现了专家级别的性能.
科学领域:
- 医疗成像医学成像
- 人工智能的人工智能
- 放射学质量保证 质量保证
背景情况:
- 乳房镜质量控制对于准确的癌症检测至关重要.
- 放射科医生间的变化可能会影响评估的一致性.
- 韩国国家认证系统要求严格的幻影影像评估.
研究的目的:
- 开发和验证一个深度学习模型用于自动化乳房镜像幻影图像评估.
- 加强放射科医生之间的协议和质量控制的效率.
- 为了使自动评估与既定指南标准保持一致.
主要方法:
- 利用了来自KIAMI的5917张乳房镜像幻影图像,其中5813张符合质量标准.
- 采用了EfficientNetV2_L深度学习模型来预测病变的存在和得分 (0.0,0.5,1.0).
- 使用精度,F1得分,AUC和可解释的AI技术评估模型性能.
主要成果:
- 对于纤维 (87.84%),斑点 (93.43%) 和质量 (86.63%) 实现了高分类准确度.
- 在所有评分级别检测病变方面表现出强的表现 (AUC>0.94为0.5分).
- 可解释的AI证实了对得分指南和病变特征的遵守.
结论:
- 深度学习模型提供精确的,专家级评估的乳房影像幻影图像.
- 自动化评估提高了评分的一致性和工作流程效率.
- 该模型可扩展,以加强国家质量控制系统.
相关概念视频
01:05Quality Control
Quality control is one of the three cyclical quality assurance activities that help keep a system under statistical control. Typical quality control activities include creating quality control charts, conducting proficiency testing, and documenting and archiving results.
Quality control helps track data, visualize trends, and identify variations, making it easier to detect deviations that may affect the accuracy of an analysis. One way to do this is by generating a quality control chart, which...
Quality control helps track data, visualize trends, and identify variations, making it easier to detect deviations that may affect the accuracy of an analysis. One way to do this is by generating a quality control chart, which...
01:19Quality Assurance
Quality assurance is the overarching term used to describe the activities employed to ensure the proper performance of a system. These activities can be classified into three categories: quality control, quality assessment, and internal corrective measures. Typically, these activities work cyclically: quality control is performed before and during the analysis, while quality assessment occurs during and after the investigation. Internal corrective measures are implemented based on the findings...