<p>The purpose of this study was to evaluate inter-scanner agreement and the dose-related stability of negative CT-number measurements on dual-energy CT virtual monoenergetic images under low-, mid-, and high-dose conditions. A custom phantom (sawdust, salad oil, cutting-fluid dilutions 100%–3%, and water) was scanned on two CT systems (Revolution CT and Discovery CT750 HD). DECT VMI (40–140&#xa0;keV) was acquired at three CTDIvol-defined dose levels with five repeats per condition; CT numbers were measured using standardized central ROIs. Negative-HU materials increased monotonically with increasing keV on both scanners. At 70&#xa0;keV, inter-scanner bias (ΔHU = Revolution − 750HD) increased as HU became more negative (<i>r</i> = 0.840), with an overall mean bias of − 3.57 HU and 95% limits of agreement of − 16.90 to 9.75 HU. The largest discrepancies were for sawdust (− 18.70 to − 20.09 HU). Water reference behavior differed by scanner and dose (Mean(|HU|) across 40–140&#xa0;keV: 0.223 HU for Revolution vs. 2.464 HU for 750HD at low dose; 750HD remained offset at mid–high dose (2.319–2.508 HU). Dose stability at 70&#xa0;keV was generally high, but 750HD showed low-dose deviations for 20% and 30% (ranges 12.13 and 12.27 HU) that diminished at mid–high dose. VMI shows predictable energy-dependent behavior for negative-HU materials, but inter-scanner bias increases for strongly negative values and low-dose conditions, and water-reference behavior can influence stability; scanner-aware interpretation and phantom-based harmonization are warranted for quantitative evaluation in the negative HU range.</p>

错误:搜索内容不能为空,请输入英文关键词
错误:关键词超出字数限制,请精简
高级检索

Intra-vendor inter-scanner agreement of negative-HU measurements on fast kV-switching dual-energy CT virtual monoenergetic imaging: a phantom study

  • Norimi Nishiyama,
  • Takanori Masuda,
  • Junnichi Nakagawa,
  • Motohiro Tabchi,
  • Ryota Yamanaka,
  • Yoshinori Nagayasu,
  • Yuta Fujiwara,
  • Hiroyuki Ikenaga

摘要

The purpose of this study was to evaluate inter-scanner agreement and the dose-related stability of negative CT-number measurements on dual-energy CT virtual monoenergetic images under low-, mid-, and high-dose conditions. A custom phantom (sawdust, salad oil, cutting-fluid dilutions 100%–3%, and water) was scanned on two CT systems (Revolution CT and Discovery CT750 HD). DECT VMI (40–140 keV) was acquired at three CTDIvol-defined dose levels with five repeats per condition; CT numbers were measured using standardized central ROIs. Negative-HU materials increased monotonically with increasing keV on both scanners. At 70 keV, inter-scanner bias (ΔHU = Revolution − 750HD) increased as HU became more negative (r = 0.840), with an overall mean bias of − 3.57 HU and 95% limits of agreement of − 16.90 to 9.75 HU. The largest discrepancies were for sawdust (− 18.70 to − 20.09 HU). Water reference behavior differed by scanner and dose (Mean(|HU|) across 40–140 keV: 0.223 HU for Revolution vs. 2.464 HU for 750HD at low dose; 750HD remained offset at mid–high dose (2.319–2.508 HU). Dose stability at 70 keV was generally high, but 750HD showed low-dose deviations for 20% and 30% (ranges 12.13 and 12.27 HU) that diminished at mid–high dose. VMI shows predictable energy-dependent behavior for negative-HU materials, but inter-scanner bias increases for strongly negative values and low-dose conditions, and water-reference behavior can influence stability; scanner-aware interpretation and phantom-based harmonization are warranted for quantitative evaluation in the negative HU range.