减少联邦学习中影像标注疲劳:客户端自适应视觉语言门控
本研究基于公开的Emory乳腺影像数据集,开发并评估一种客户端自适应的视觉语言门控模型PromptGate,用于在联邦学习乳腺密度标注任务中过滤含放射伪影的不相关图像。回顾性研究显示,该模型对伪影的敏感度为65.6%,ID图像保留率84.4%,标注时间减少约30%,队列纯度达92.5%,下游密度准确率与未过滤时相似。
要点
- PromptGate可自动拦截65.6%的伪影图像,将标注队列纯度从88.5%提升至92.5%。
- 过滤后标注总时间减少约30%,实际获益受各扫描仪伪影患病率影响。
- 该技术仍会错误剔除15.6%的正常ID图像,需前瞻性多中心验证后再部署。
核心结论
- 客户端自适应视觉语言门控可在联邦乳腺密度标注中提高任务相关图像浓度,减少标注阅读总时间。
- 部分ID图像被错误剔除,且本研究使用伪影富集队列,需前瞻性多读片者、多中心验证后方可临床部署。
研究设计与样本
- 回顾性、按扫描仪划分的研究,基于公开Emory乳腺影像数据集FedEMBED(训练222700张,测试65891张,4台扫描仪,各扫描仪伪影患病率4.8%–13.1%)。
- PromptGate在生物医学视觉语言模型上适配站点特异性可学习提示,以softmax概率最高类别将图像分为相关ID或不相关OOD。
- 1名具有2年乳腺影像经验的研究团队成员独立阅读患者分离的500张伪影富集单视图(ID 346张,OOD 154张),对参考标签、门控决策和扫描仪信息设盲。
关键影像表现/征象
- PromptGate从154张含伪影图像中拦截101张,从346张ID图像中错误剔除54张。
- 伪影判别ROC曲线下面积(AUC)为0.85。
- 读片者与手动伪影参考的一致性为93.6%(κ=0.85),与临床记录密度标签的精确一致性为60.9%(κ=0.48)。
- 队列纯度达92.5%,无过滤时为88.5%,静态视觉语言模型过滤时为87.0%。
诊断标准与关键数值
- 伪影敏感度65.6%,ID保留率84.4%,ID图像中假剔除率15.6%。
- 富集队列中标注总时间从3.82小时降至2.68小时,减少1.14小时(约30%);仅计入正确剔除伪影的估计节省20.0%。
- 基于扫描仪特定患病率外推,每500张图像可节省约20–65分钟,但绝对获益在富集队列中较小。
- 下游乳腺密度准确率与未过滤条件相似(原文未给出具体数值)。
临床应用与阅片要点
- 日常乳腺密度标注时,可先行自动过滤含放射伪影的视图,提高队列中有效图像的浓度,减少无效阅片。
- 不同扫描仪伪影患病率差异较大,部署时应按设备校准过滤比例,避免过度剔除正常图像。
- 下游密度判读准确率不因过滤而显著改变,说明门控在队列层面节省时间而不损害任务性能。
局限性
- 单中心回顾性研究,读片者仅1名且为研究团队成员,缺乏多读片者外部验证。
- 研究队列为伪影富集设计,标注时间节省的绝对获益预计小于富集队列观测值。
- 15.6%的ID图像被错误剔除,可能导致有效数据丢失。
以上解读依据论文摘要(全文未开放获取或许可不允许翻译转载)。
英文原文(摘要)· 点击展开
PURPOSE: Federated learning enables breast-imaging sites to jointly train mammography artificial intelligence (AI) without sharing images, but radiologists at each site must still annotate selected images during active-learning rounds. We developed and evaluated a client-adaptive vision-language gatekeeper that withholds radiopaque artifact-containing images considered unsuitable for a breast-density annotation objective and quantified its effect on annotation time, queue quality, and label reliability.
METHODS: In this retrospective, scanner-partitioned study, we used FedEMBED, derived from the public Emory Breast Imaging Dataset (222,700 training and 65,891 test images; four scanners; artifact prevalence 4.8%-13.1% per scanner). PromptGate adapted a biomedical vision-language model with site-specific learnable prompts and classified each image as relevant in-distribution (ID) or irrelevant out-of-distribution (OOD) using the class with the highest softmax probability. A single radiologist who is a member of the study team and has 2 years of breast imaging experience independently read a patient-separated, artifact-enriched cohort of 500 single mammographic views (ID, n=346; OOD, n=154). Cases were randomized, ID and OOD images were intermixed, and the reader was blinded to reference labels, PromptGate decisions, enrichment status, and scanner metadata.
RESULTS: PromptGate withheld 101 of 154 artifact-containing images and 54 of 346 ID images, corresponding to 65.6% artifact sensitivity, 84.4% ID retention, and a 15.6% false-removal rate among ID images. Gross annotation time decreased by ∼30% (3.82 to 2.68 hours; 1.14 hours) in the enriched cohort; an analysis crediting only correctly removed artifacts yielded an estimated 20.0% reduction. A simple prevalence-based extrapolation suggests an approximate 20- to 65-min saving per 500 images at the observed scanner-specific prevalence, although the absolute benefit is expected to be smaller than in the enriched cohort. Artifact discrimination reached an area under the receiver operating characteristic curve of 0.85. Reader agreement with the manual artifact reference was 93.6% (κ=0.85), whereas agreement with the clinical-record density label was 60.9% exact (κ=0.48). Queue purity reached 92.5%, compared with 88.5% without filtering and 87.0% with a static vision-language model filter, while downstream density accuracy remained similar to the unfiltered condition.
CONCLUSION: In this retrospective proof-of-concept study, a privacy-oriented, scanner-adaptive vision-language gate improved the concentration of task-relevant images in federated breast-density annotation queues and reduced gross reading time. Because some ID images were incorrectly withheld and the reader study used an artifact-enriched cohort, a prospective multireader and multi-institutional evaluation is needed before operational deployment.
引用
- Nesturi, Gaviria, Zeng, et al. Reducing Clinician Annotation Fatigue in Open-Set Federated Learning: A Client-Adaptive Vision-Language Gatekeeper. J Am Coll Radiol. 2026. PMID: 42822685. doi:10.1016/j.jacr.2026.09.023
供执业医师工作参考,不替代临床判断;数值与分级以指南原文为准。