文献速递心脏核医学

减少光子计数水平及正常数据选择对心脏SPECT半自动图像评估的影响

Effect of reduced photon count levels and choice of normal data on semi-automated image assessment in cardiac SPECT.
PubMed 文献 · J Nucl Cardiol · 2018-04-13

研究纳入50例接受Tc-99m甲氧基异丁基异腈一日法负荷/静息SPECT/CT的患者,用二项稀疏法生成多级计数降低,以Corridor 4DM分别采用计数匹配或全计数正常数据计算SSS,并将研究分为低危(SSS<4)与中/高危(SSS≥4)。计数匹配正常数据下计数降低增加假正常率并降低灵敏度;全计数正常数据下计数降低增加假低灌注率并降低特异度;约67%剂量及以下时重分类率显著。

要点

  • 计数匹配正常数据时,计数降低增加假正常率并降低灵敏度。
  • 全计数正常数据时,计数降低增加假低灌注率并降低特异度。
  • 约67%剂量及以下时,重分类率显著。

核心结论

  • 正常数据相对于实际患者数据的计数水平会导致显著偏倚。
  • 与标准LEHR相比,IQ-SPECT应允许显著降低剂量。

研究设计与样本

  • 50例患者接受Tc-99m甲氧基异丁基异腈一日法负荷/静息SPECT/CT。
  • 使用二项稀疏法生成多级计数降低。
  • 使用Corridor 4DM,分别以计数匹配或全计数正常数据计算SSS。
  • 研究分为低危(SSS<4)或中/高危(SSS≥4)。

主要发现

  • 计数匹配正常数据下计数降低增加假正常率并降低灵敏度。
  • 全计数正常数据下计数降低增加假低灌注率并降低特异度。
  • 约67%剂量及以下时重分类率显著。

关键数值与推荐

  • 低危定义为SSS<4,中/高危定义为SSS≥4。
  • 重分类率在约67%剂量及以下时显著。

临床怎么用

  • 原文未据此给出推荐。

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)。

原文

BACKGROUND: The SMARTZOOM multifocal collimator from Siemens Healthcare was developed to improve the γ-photon sensitivity in myocardial perfusion imaging without truncating the field of view. As part of the IQ-SPECT package, it may be used to reduce radiopharmaceutical dose to patients, as well as acquisition time. The aim of this study was twofold: (1) to evaluate the influence of dose reduction in semi-automated MPI scoring, with focus on different strategies for the choice of normal data (count-matched, full-count), and (2) to evaluate the effect of dose reduction afforded by Siemens' IQ-SPECT package. METHODS: 50 patients underwent Tc-99m-sestamibi one-day stress/rest SPECT/CT. Multiple levels of count reduction were generated using binomial thinning. Using Corridor 4DM, summed stress score (SSS) was calculated using either count-matched or full-count normal data. Studies were classified as low-risk (SSS < 4) or intermediate/high-risk (SSS ≥ 4). RESULTS: Count reduction using count-matched normal data increases false-normal rate and decreases sensitivity. With full-count normal data, count reduction increases false-hypoperfusion rate, leading to decreased specificity. Altogether, rate of reclassification was significant at roughly 67% dose and below. CONCLUSION: Significant bias results from count level of normal data relative to actual patient data. Compared to standard LEHR, IQ-SPECT should allow for significant dose reduction.

来源:J Nucl Cardiol,查看原文。版权归原作者所有。

引用

  1. Wetzl M, Sanders JC, Kuwert T, et al. Effect of reduced photon count levels and choice of normal data on semi-automated image assessment in cardiac SPECT. J Nucl Cardiol. 2018. PMID: 29654444. doi:10.1007/s12350-018-1272-1

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。