文献速递腹部MR

超分辨率深度学习重建2D MRCP改善主胰管显示

Super-resolution deep learning-enhanced 2D MRCP improves main pancreatic duct visualization compared with 3D MRCP.
PubMed 文献 · Eur J Radiol · 2026-06-16

本研究为前瞻性队列研究,纳入174例接受MRCP胰腺评估的患者,比较3D MRCP联合超分辨率深度学习重建(3D-SR)、传统压缩感知2D MRCP(2D-CS)及超分辨率深度学习重建2D MRCP(2D-SR)。结果显示2D-SR对主胰管总体显示最佳,3D-SR在远端胆总管定量指标最高。2D技术对无上游扩张的局灶性主胰管狭窄假阳性率更低。

要点

  • 2D-SR对主胰管体尾部显示与边缘锐利度优于2D-CS和3D-SR。
  • 3D-SR假阳性率较高,提示发现局灶性主胰管狭窄时需谨慎。
  • 对≤5mm囊性病灶,2D技术显示清晰度优于3D-SR。

核心结论

  • 在该监测队列中,2D-SR对主胰管总体显示最佳。
  • 2D技术对无上游扩张的局灶性主胰管狭窄显示更低的假阳性率,但该发现为探索性。
  • 3D-SR对远端胆总管SNR、CR和CNR均最高。

研究设计与样本

  • 前瞻性研究,纳入174例接受MRCP胰腺评估的患者。
  • 比较3D-SR、2D-CS和2D-SR三种MRCP重建技术。
  • 由两名放射科医师评估图像质量、主胰管显示、囊性病灶清晰度和主胰管异常。

关键影像表现/征象

  • 主胰管体尾部显示和边缘锐利度方面,2D-SR最高,其次为2D-CS,3D-SR最低。
  • 囊性病灶清晰度方面,2D-SR和2D-CS高于3D-SR,且≤5mm病灶差异最大。
  • 三种技术均正确识别伴上游扩张的局灶性主胰管狭窄3例和弥漫性主胰管扩张13例。
  • 对无上游扩张的局灶性狭窄3例,3D-SR假阳性率为40.4%至42.7%,两种2D技术均无误判。

诊断标准与关键数值

  • 3D-SR对远端胆总管的SNR、CR和CNR均为三组最高(所有p≤0.006)。
  • 总体图像质量、运动伪影评分、主胰管体尾部显示和边缘锐利度方面,2D-SR和2D-CS均高于3D-SR,且2D-SR高于2D-CS(所有p≤0.037)。
  • 对无上游扩张的局灶性主胰管狭窄,3D-SR假阳性率为40.4%至42.7%,2D技术假阳性率为0。

临床应用与阅片要点

  • 需要评价主胰管体尾部及细微狭窄时,可优先选择2D-SR或2D-CS序列。
  • 3D-SR对局灶性主胰管狭窄易产生假阳性,阅片时需结合其他序列确认。
  • 观察≤5mm胰腺囊性病灶时,2D技术可能提供更好的病灶清晰度。

局限性

  • 对无上游扩张的局灶性主胰管狭窄仅3例真阳性,相关假阳性率差异仅为探索性。
  • 研究队列为监测人群,可能不能代表所有胰腺疾病谱。

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: The relative merits of 2D and 3D MRCP remain incompletely defined, and the effect of super-resolution deep learning reconstruction (SR-DLR) on main pancreatic duct (MPD) visualization remains unclear. We compared image quality and MPD visualization among 3D MRCP with SR-DLR (3D-SR), conventional 2D MRCP with compressed sensing (2D-CS), and 2D MRCP with SR-DLR (2D-SR), and exploratorily assessed diagnostic performance for MPD abnormalities.

MATERIALS AND METHODS: This prospective study included 174 patients undergoing MRCP for pancreatic evaluation. Quantitative metrics included signal-to-noise ratio (SNR), contrast ratio (CR), and contrast-to-noise ratio (CNR) for the distal common bile duct and pancreatic cystic lesions. Two radiologists assessed image quality, MPD visualization, cystic lesion conspicuity, and MPD abnormalities.

RESULTS: 3D-SR yielded the highest SNR, CR, and CNR for the distal common bile duct (all p ≤ 0.006). Overall image quality, motion artifact scores, MPD body/tail visualization, and margin sharpness were higher with 2D-SR and 2D-CS than 3D-SR, and higher with 2D-SR than 2D-CS (all p ≤ 0.037). Cystic lesion conspicuity was higher with 2D-SR and 2D-CS than 3D-SR, with the greatest difference for lesions ≤ 5 mm. All techniques correctly identified focal MPD narrowing with upstream dilatation (n = 3) and diffuse MPD dilatation (n = 13). For focal narrowing without upstream dilatation (n = 3), 3D-SR showed a higher false-positive rate (readers 1-2: 40.4-42.7%), whereas neither 2D technique showed false-positive findings.

CONCLUSION: In this surveillance cohort, 2D-SR provided the best overall MPD visualization; the lower false-positive rate for focal MPD narrowing with 2D techniques (only three true-positive cases) was exploratory and hypothesis-generating.

来源:Eur J Radiol,查看原文。版权归原作者所有。

引用

  1. Ozaki, Haseyama, Sugioka, et al. Super-resolution deep learning-enhanced 2D MRCP improves main pancreatic duct visualization compared with 3D MRCP. Eur J Radiol. 2026. PMID: 42320143. doi:10.1016/j.ejrad.2026.113022

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。