动态AI辅助同侧组织匹配用于数字乳腺断层摄影
研究分两部分:14名放射科医生对11例主观评价AI工具,12名医生在30例94个共识参考标注上比较有无AI辅助的病灶标注误差。无AI时异常病灶RMSE高32%(11.70 mm vs 8.88 mm)、MDE高37.5%(20.68 mm vs 15.08 mm),非专家RMSE与MDE降低超60%。
要点
- 无AI辅助时异常病灶定位误差RMSE为11.70 mm,有AI辅助时为8.88 mm。
- 非专家阅片者受益最大,RMSE 12.20 mm vs 7.57 mm、MDE 15.76 mm vs 9.47 mm。
- AI警告系统可有效识别计算映射高度不确定的病例。
核心结论
- AI辅助组织匹配显著提高定位准确性,尤其对非专家阅片者和复杂病例;AI警告系统有效识别高不确定性病例,误差降至典型肿瘤尺寸范围内,可能避免漏诊病灶。
研究设计与样本
- 研究分两部分:第1部分由14名放射科医生对11例主观评价AI工具对信心和感知有用性的影响;第2部分由12名放射科医生在洗脱期后分别在有、无AI辅助下标注病灶,以两名专家审阅者建立的30例94个共识参考标注为参照,按专业水平、病灶异常类型和ViewFinder警告分层,采用Wilcoxon符号秩检验。
诊断标准与关键数值
- 无AI辅助时异常病灶RMSE为11.70 mm,有AI辅助时为8.88 mm,无AI辅助高32%(p<0.05)。
- 无AI辅助时异常病灶MDE为20.68 mm,有AI辅助时为15.08 mm,无AI辅助高37.5%(p<0.05)。
- 非专家阅片者无AI辅助时RMSE为12.20 mm,有AI辅助时为7.57 mm(p<0.01)。
- 非专家阅片者无AI辅助时MDE为15.76 mm,有AI辅助时为9.47 mm(p<0.05),降低超60%。
- 第1部分平均有用性评分为6.21/10(p<0.001)。
临床应用与阅片要点
- AI辅助组织匹配可帮助放射科医生,尤其非专家,将病灶定位误差缩小至典型筛查检出乳腺肿瘤尺寸(中位数13 mm,IQR 9-20 mm)范围内。
- ViewFinder警告提示计算映射高度不确定,阅片时可据此识别高不确定性病例。
局限性
- 原文未报告局限性
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: To evaluate whether AI-assisted ipsilateral tissue matching in digital breast tomosynthesis (DBT) reduces localization errors beyond typical tumor boundaries, particularly for non-expert radiologists. The technology category is deep learning.
MATERIALS AND METHODS: The study consisted of two parts. In Part 1, 14 radiologists subjectively evaluated the AI tool's impact on confidence and perceived usefulness on 11 cases. In Part 2, 12 radiologists marked lesion annotations with and without AI assistance after a washout period. Errors were measured relative to 94 consensus reference annotations in 30 cases established by two expert reviewers. Stratifications included expertise, lesion abnormality, and presence of ViewFinder warnings (indicating high uncertainty in the computed mapping). Wilcoxon signed-rank tests assessed statistical significance.
RESULTS: In Part 1, radiologists reported improved confidence and a mean usefulness score of 6.21/10 (p < 0.001), with greater benefit in challenging cases. In Part 2, root mean square error (RMSE) and maximum distance error (MDE) for abnormal lesions were 32 % (11.70 mm vs 8.88 mm) and 37.5 % higher (20.68 mm vs 15.08 mm), respectively, without AI assistance (p < 0.05). Non-expert readers showed the greatest benefit, with RMSE (12.20 mm vs 7.57 mm, p < 0.01) and MDE (15.76 mm vs 9.47 mm, p < 0.05) reduced by over 60 %. These reductions are clinically relevant given typical screen-detected breast tumor sizes (median, 13 mm [IQR: 9-20 mm]).
CONCLUSION: AI-assisted tissue matching significantly improved localization accuracy, especially for non- expert readers and in complex cases. The AI warning system effectively identified high-uncertainty cases. Errors were reduced to within typical tumor dimensions, potentially preventing missed lesions.
引用
- Morrell S, Hutel M, Lucena O, et al. Dynamic AI-assisted ipsilateral tissue matching for digital breast tomosynthesis. Eur J Radiol. 2025. PMID: 40925265. doi:10.1016/j.ejrad.2025.112356
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。