文献速递胸部技术与安全CT

光子计数CT超低剂量与低剂量肺定量定性参数比较

Quantitative and qualitative lung parameters in photon-counting CT: comparison of ultra-low-dose with standard low-dose protocols.
PubMed 文献 · Eur Radiol · 2026-09-18

本前瞻性研究纳入101例患者,同次接受低剂量与超低剂量光子计数CT胸部扫描,比较定量与定性肺参数。超低剂量使噪声升高、外周气道分割减少、基于密度的肺气肿定量出现偏倚且视觉肺气肿评分低估。结论支持剂量分层解读,超低剂量定量结果需谨慎用于临床。

要点

  • 超低剂量光子计数CT会低估视觉肺气肿程度,报告时需注意剂量条件。
  • 超低剂量下气道分割分支数减少,外周气道评估可靠性下降。
  • 基于密度的肺气肿指数在超低剂量时偏高,不能直接与低剂量结果混用。

核心结论

  • 超低剂量光子计数CT可实现低剂量定量肺成像,但会降低外周气道可见性与分割完整性,并使基于密度的肺气肿定量及视觉评估产生偏倚。
  • 超低剂量来源的定量CT参数需在剂量校正策略验证后方可常规临床应用。

研究设计与样本

  • 前瞻性研究,101例患者同次接受低剂量(CTDIvol 0.64±0.15 mGy)与超低剂量(CTDIvol 0.11±0.03 mGy)光子计数CT胸部扫描。
  • 自动化定量CT在软组织核(Br36u)与肺核(Bl64u)上评估噪声指数、气道分割、气道壁百分比、肺气肿指数、平均肺密度与Percentile15。
  • 两名阅片者采用Likert量表对图像质量、气道及肺气肿范围进行评分。

关键影像表现/征象

  • 超低剂量较低剂量噪声升高:肺核Bl64u增加94.4%,软组织核Br36u增加32.8%。
  • 超低剂量自动化气道分割分支数减少(平均减少41.4支,p<0.001),可见气道级数减少(p<0.001)。
  • 视觉支气管壁增厚在两名阅片者均无显著差异(R1 p=0.89;R2 p=0.20)。
  • 超低剂量视觉肺气肿评估向较轻程度偏移(p<0.001)。

诊断标准与关键数值

  • 气道壁百分比在Bl64u上超低剂量高于低剂量(LD-ULD -3.33%;p<0.001),外周差异更大;Br36u全树值无差异(LD-ULD -0.05%;p=0.83)。
  • 超低剂量平均肺密度较低剂量低12.5 HU(p<0.001),Percentile15低19.9 HU(p<0.001),肺气肿指数高6.2%。
  • 视觉图像质量超低剂量显著低于低剂量(p<0.001)。

临床应用与阅片要点

  • 阅片时应根据剂量条件判断肺气肿视觉评分,超低剂量图像可能低估肺气肿严重程度。
  • 超低剂量下外周气道分支显示减少,报告外周气道病变时需考虑分割与可见性限制。
  • 不同重建核下气道壁百分比结果不一致,对比定量气道参数时需固定核与气道级别。

局限性

  • 研究为单中心前瞻性设计,样本量为101例,混合适应证,未报告多中心验证。
  • 原文指出超低剂量定量CT需进一步验证剂量校正策略后方可常规临床应用。

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: Photon-counting CT (PCCT) enables dose reduction beyond energy-integrating CT, but ultra-low-dose acquisitions may impair assessment of airways and lung parenchyma. We evaluated the impact of ultra-low-dose (ULD) PCCT on quantitative and qualitative lung parameters.

METHODS: In this prospective study, 101 patients underwent same-session low-dose (LD) and ULD chest CT for mixed indications (CTDIvol, 0.64 ± 0.15 vs. 0.11 ± 0.03 mGy). Automated QCT assessed noise index, airway segmentation, airway wall percentage, emphysema index (EI), mean lung density (MLD), and Percentile15 (Perc15) on soft (Br36u) and lung (Bl64u) kernels. Two readers scored image quality, airways, and emphysema extent using Likert scales.

RESULTS: ULD increased noise compared with LD (Bl64u: +94.4%; Br36u: +32.8%) and reduced visual image quality (p < 0.001). Automated airway segmentation decreased (-41.4 mean no. of branches, p < 0.001), with fewer visible airway generations (p < 0.001). Visual bronchial wall thickening did not differ (R1 p = 0.89; R2 p = 0.20). Airway wall percentage showed kernel- and generation-dependent differences: whole-tree values did not differ on Br36u (LD-ULD -0.05%; p = 0.83) but were higher at ULD on Bl64u (LD-ULD -3.33%; p < 0.001), with greater differences peripherally. At ULD, MLD and Perc15 were lower than at LD by 12.5 and 19.9 HU, respectively (both p < 0.001), while EI was 6.2% higher. Visual emphysema assessment shifted toward lower severity at ULD (p < 0.001).

CONCLUSION: ULD-PCCT enables dose-efficient QCT but introduces limitations: reduced peripheral airway visibility/segmentation with differences in airway wall measurements, biased density-based emphysema quantification, and visual emphysema underestimation at ULD, supporting dose-aware interpretation and further validation of correction strategies.

KEY POINTS: Question To study the influence of ultra-low-dose photon-counting chest CT on quantitative and qualitative airway and emphysema assessment compared to low-dose CT. Findings Ultra-low-dose increased noise and reduced peripheral airway segmentation; airway wall metrics showed kernel- and generation-dependent differences, with biased density-based and visual emphysema assessment. Clinical relevance PCCT enables quantitative lung imaging at ultra-low dose, but dose reduction alters airway segmentation and density-based emphysema metrics. ULD-derived QCT therefore requires cautious interpretation and validation of dose-aware correction strategies before routine clinical use.

来源:Eur Radiol,查看原文。版权归原作者所有。

引用

  1. Kroschke, Jungblut, von Stackelberg, et al. Quantitative and qualitative lung parameters in photon-counting CT: comparison of ultra-low-dose with standard low-dose protocols. Eur Radiol. 2026. PMID: 42760429. doi:10.1007/s00330-026-12887-9

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。