In 51 patients with 51 coronary plaques, PCCT VMI energy had a substantial effect on measured plaque attenuation and consequently on quantitative plaque composition when fixed HU thresholds were used. Mean plaque attenuation fell progressively from 723 ± 501 HU at 40 keV to 120 ± 112 HU at 180 keV, while 70-keV VMI most closely reproduced the 120-kVp T3D reference attenuation.
The effect differed by plaque component. With the primary threshold method, calcified plaque volume decreased from 138.7 ± 126.4 mm³ at 40 keV to 38.5 ± 64.6 mm³ at 180 keV, whereas low-attenuation plaque increased from 22.8 ± 24.9 to 96.0 ± 76.3 mm³. Thus, a plaque can appear quantitatively very different solely because the VMI energy changes.
Low-keV images provided the highest CNR and SNR but also more noise. The key practical implication is that fixed plaque-attenuation thresholds developed for conventional images should not simply be transferred across PCCT VMI energies; standardized energy-specific thresholds or normalization strategies are needed for reproducible plaque quantification.
Access publication
Figure1. Representative CTA images of coronary plaques reconstructed in T3D and different VMI energy levels (40, 70, 120, and 180 keV). Quantitative plaque analyses of a partially calcified-predominantly non-calcified (panel A) and partially calcified-predominantly calcified (panel B) plaque are shown in T3D and different VMI reconstructions at 40, 70, 120, and 180 keV levels. The red line illustrates the border of the vessel wall and the orange line illustrates the lumen border segmented on T3D images. Corresponding cross-sectional images are also depicted at the point of the maximal narrowing of the lesion. The same window setting was applied for all represented images: window: 800; level: 250. Abbreviations: CP, calcified plaque; HU, Hounsfield unit; LAP, low-attenuation non-calcified plaque; NCP, non-calcified plaque.
Protocol
Detail | Value |
|---|
Scanner | NAEOTOM Alpha |
Scan area | Coronary CTA |
Acquisition mode | ECG-triggered PCCT |
Scan mode | TurboFlash high-pitch, sequential, or helical depending on heart rate/rhythm |
Tube voltage [kV] | 120 |
IQ level | 80 |
Dose modulation | Automatic tube current modulation |
Rotation time [s] | 0.25 |
Detector configuration | 144 × 0.4 mm |
Slice thickness [mm] | 0.4 |
Reconstruction increment [mm] | 0.4 |
Matrix | 512 × 512 |
Kernel | Bv40 |
Iterative reconstruction | QIR 2 |
Reference reconstruction | 120-kVp polychromatic T3D |
VMI energies | 40–180 keV in 10-keV increments |
ECG phase | Diastolic 65–85% RR or systolic 200–400 ms depending on heart rate |
Contrast volume | 70–80 mL |
Injection rate | 4.5–5.0 mL/s |
Contrast protocol | Four-phasic |
Heart-rate control | IV beta-blocker if HR >65 bpm |
Vasodilation | 0.8 mg sublingual nitroglycerin if systolic BP >100 mmHg |
Plaque analysis software | AutoPlaque 2.5 |
Primary plaque thresholds | LAP −100 to 30 HU; NCP 30–350 HU; CP >350 HU |
Key Results
Parameter | Main finding |
|---|
Study population | 51 patients / 51 plaques |
Plaque attenuation | 723 ± 501 HU at 40 keV → 120 ± 112 HU at 180 keV |
CNR | Highest at 40 keV: 22.1 ± 5.6; 70 keV similar to T3D |
Calcified plaque volume | 138.7 ± 126.4 mm³ at 40 keV → 38.5 ± 64.6 mm³ at 180 keV |
Low-attenuation plaque volume | 22.8 ± 24.9 mm³ at 40 keV → 96.0 ± 76.3 mm³ at 180 keV |
Closest VMI to T3D | 70 keV for overall attenuation/CP; 50 keV for LAP with primary thresholds |
Source:
Vattay B, Szilveszter B, Boussoussou M, Vecsey-Nagy M, Lin A, Konkoly G, Kubovje A, Schwarz F, Merkely B, Maurovich-Horvat P, Williams MC, Dey D, Kolossváry M. Impact of virtual monoenergetic levels on coronary plaque volume components using photon-counting computed tomography. European Radiology. 2023;33:8528–8539. doi:10.1007/s00330-023-09876-7.
