6.8%的脑部MRI重建研究同时评估了图像保真度和医生读片，暴露了当前评价体系的重大盲区。
Evaluating the Safety of Deep Learning-Based Brain MRI Reconstruction
- 通过系统综述整合263项研究，分析现有评估方法
- 仅6.8%的研究同步进行图像质量与医生读片评估
- 生成模型幻觉问题严重但极少被医生验证
目的：深度学习使脑部MRI加速4至10倍，但模型可能抹除病灶或合成虚假组织——这些错误在像素级指标如PSNR和SSIM下难以察觉。我们评估当前评价实践是否能发现这一盲点。方法：遵循PRISMA 2020，无时间限制地检索七个数据库，共纳入263项研究（1995–2026），使用QUADAS-2评估质量并匹配工具，进行叙述性综合。分类基于标题、摘要及受控词汇；报告的流行率代表下限。重复筛选达成高一致性（Fleiss kappa = 0.877），评估一致性为0.788（可观测部分为0.390）。数据提取未经审计。结果：仅18项研究（6.8%）在同一数据上同时记录了保真度指标和医生读片评估，核心替代指标未被测量。医生研究多关注组间一致性，结果较弱：fastMRI 2020中一致性分别为0.457和0.386（Kendall W），仅在SSIM偏离时改善。在所述误差模型下，抹除100 mm³腔隙性梗死仅使全局PSNR变化0.03 dB。随着文献量增长五倍，医生评估比例从32%降至18%，后回升至21%。生成模型（最常关联幻觉，占比39%）是医生评估最少的（11.3%），而自监督模型达47%且零医生评估。仅5%发布代码并开展医生评估；无人评估模型观察者；无命名数据集涵盖急性卒中或出血。结论：基于这些下限，当前评价实践无法保证诊断安全性。我们提出五项面向安全的评估必须满足的要求。
原文摘要 · Abstract (English)
Objective: Deep learning accelerates brain MRI four- to tenfold, but models can erase lesions or synthesize false tissue - failures pixel-averaged metrics like PSNR and SSIM miss. We review whether current evaluation practices detect this blind spot. Methods: Following PRISMA 2020, we searched seven databases without date limits, including 263 studies (1995-2026), appraised them using QUADAS-2 and matched instruments, and synthesized narratively. Categories were derived from titles, abstracts, and controlled vocabulary; reported prevalence figures represent floors. Duplicate screening achieved high agreement (Fleiss kappa = 0.877), as did appraisal (0.788; 0.390 where observable). Extraction is unaudited. Results: Only 18 of 263 studies (6.8%) recorded both a fidelity metric and reader assessment on identical data, leaving the central surrogate unmeasured. Reader studies mostly measured inter-reader agreement, which was weak: fastMRI 2020 concordance reached 0.457 and 0.386 (Kendall W), improving only where SSIM diverged. Erasing a 100 mm3 lacunar infarct shifts global PSNR by 0.03 dB under the stated error model. As the corpus grew fivefold, reader assessments dropped from 32% to 18%, recovering to 21%. Generative models - most associated with hallucination (39%) - were among the least reader-evaluated (11.3%), while self-supervised models reached 47% with zero reader evaluation. Only 5% released code and ran reader studies; none evaluated a model observer; no named dataset covered acute stroke or hemorrhage. Conclusions: On these floors, current evaluation practices cannot certify diagnostic safety. We derive five requirements safety-oriented evaluations must meet.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。