文献速递腹部CT

双期深度学习提高对比增强CT上偶发小胰腺囊性病变(0.5-3 cm)的检出

Dual-Phase deep learning Enhances detection of incidental small pancreatic cystic lesion (0.5-3 cm) on Contrast-Enhanced CT.
PubMed 文献 · Eur J Radiol · 2025-11-07

回顾性收集437例偶发小胰腺囊性病变(含201例亚厘米囊肿)及193例正常胰腺的对比增强CT图像,构建动脉期与门静脉期双期深度学习模型,并与门静脉单期模型及高、低年资放射科医师比较检出效能。双期模型敏感度优于单期模型及低年资医师,与高年资医师相当,辅助后可提高放射科医师敏感度并缩短低年资医师检出时间。

要点

  • 双期深度学习模型对偶发小胰腺囊性病变的检出敏感度优于门静脉单期模型(验证队列91.7% vs 82.1%)。
  • 在测试队列中双期模型敏感度和准确率优于低年资医师,与高年资医师相当。
  • 双期模型识别出随访中进展的全部18例偶发胰腺囊性病变(其中3例恶性)。

核心结论

  • 双期深度学习模型对小胰腺囊性病变的检出性能优于门静脉单期模型和低年资放射科医师,且表现更稳健。
  • 该模型可能缩小不同经验水平医师之间的检出差距,并有助于高危病变的早期诊断。

研究设计与样本

  • 回顾性收集2018年1月至2020年12月的对比增强CT图像,共437例偶发小胰腺囊性病变(其中201例为0.5-1 cm亚厘米囊肿)和193例正常胰腺。
  • 数据随机分为训练、验证和测试队列。
  • 以双期深度学习模型(DPDL)为研究对象,与门静脉单期深度学习模型(SPDL)及高、低年资放射科医师在测试队列中比较。

诊断标准与关键数值

  • 在验证队列中,DPDL模型敏感度91.7% vs SPDL模型82.1%(P=0.021)。
  • 在测试队列中,DPDL模型敏感度92.7% vs 低年资放射科医师74.0%(P<0.001),准确率86.2% vs 69.7%(P=0.003);与高年资放射科医师比较所有P>0.05。
  • 亚组分析证实DPDL对亚厘米胰腺囊性病变的检出优于低年资放射科医师。
  • 在DPDL辅助下,放射科医师的敏感度显著提高,低年资放射科医师的检出时间和任何假阳性率(AFPR)显著降低(所有P<0.05)。
  • DPDL模型识别出随访中进展的全部18例偶发胰腺囊性病变(其中3例恶性),而低年资放射科医师在测试队列中漏诊2例。

临床应用与阅片要点

  • 双期深度学习模型可辅助检出对比增强CT上的偶发小胰腺囊性病变,尤其对亚厘米病变和低年资医师有帮助。
  • 模型辅助可提高放射科医师检出敏感度,并缩短低年资医师的检出时间、降低任何假阳性率。

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

OBJECTIVES: Develop a dual-phase deep learning (DPDL) model using arterial/portal-phase CT to detect incidental small (0.5-3 cm) pancreatic cystic lesions (PCLs).

MATERIALS: Contrast-enhanced CT images of 437 incidental small PCLs, including 201 subcentimeter cysts (0.5-1 cm) and 193 normal pancreases were retrospectively collected (January 2018 - December 2020) and randomly divided into training, validation and testing cohorts. Detection sensitivity, specificity, any-false-positive rate (AFPR) and accuracy of the DPDL model were compared with a portal single-phase deep learning (SPDL) model and the senior and junior radiologists in the testing cohort. Factors potentially affecting detection were analyzed using logistic regression analysis.

RESULTS: In the validation cohort, the DPDL model exceeded the SPDL model in sensitivity (91.7 % vs. 82.1 %; P = 0.021). In the testing cohort, it surpassed the junior radiologist in sensitivity (92.7 % vs. 74.0 %; P < 0.001) and accuracy (86.2 % vs. 69.7 %; P = 0.003), while performing comparably to the senior radiologist (all P > 0.05). Subgroup analysis confirmed DPDL's superiority for subcentimeter PCLs than the junior radiologist. With DPDL assistance, sensitivity of radiologists was significantly improved, while detection time and AFPR of the junior radiologist were significantly reduced (all P < 0.05). None of study factors affected DPDL's performance, whereas SPDL and radiologists were influenced by multiple factors. Notably, DPDL model identified all 18 incidental PCLs that progressed during follow-up (3 malignant) while the junior radiologist missed 2 in testing cohort.

CONCLUSION: The DPDL model exhibited superior and more robust detection performance for small PCLs than the SPDL model and the junior radiologist, potentially narrowing performance gaps between experience levels and improving early diagnosis of high-risk lesions.

来源:Eur J Radiol,查看原文。版权归原作者所有。

引用

  1. Deng W, Cong F, Xie F, et al. Dual-Phase deep learning Enhances detection of incidental small pancreatic cystic lesion (0.5-3 cm) on Contrast-Enhanced CT. Eur J Radiol. 2025. PMID: 41264978. doi:10.1016/j.ejrad.2025.112510

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。