In 27 patients undergoing same-day PCCT and cardiac MRI, myocardial extracellular volume (ECV) measurements remained robust across a wide range of reconstruction settings, but reconstruction choices meaningfully affected agreement with MRI.
The greatest improvement came from combining thin slices with low-keV VMI, rather than from changing iterative reconstruction strength alone. Slice thickness was a particularly important determinant of ECV accuracy. Reconstruction at 0.4/0.3 mm thickness/increment achieved the highest agreement with MRI (CCC ≈ 0.94) and substantially reduced bias compared with thicker reconstructions, likely by limiting partial-volume averaging within the myocardium.
The optimal combination was QIR 4, 0.4-mm slices with 0.3-mm increment, and 45-keV VMI. Compared with the literature-standard reconstruction of QIR 3, 1.5 mm, and 65 keV, this reduced mean ECV bias by 63% and increased concordance with MRI by 6%, bringing PCCT-derived ECV substantially closer to the MRI reference.
Access publication
Figure 1. Representation of the systematic step-by-step protocol optimization approach. In Step 1, the optimal quantum iterative reconstruction level (14) was identified. In Step 2, the different slice thicknesses were tested from 0.4 to 8 mm. In Step 3, the virtual monoenergetic image reconstruction level of the late iodine enhancement scan was optimized from 40 to 90 keV. The parameter with the best diagnostic performance in each step was retained in the subsequent step.
Protocol
Detail | Value |
|---|
Scanner | NAEOTOM Alpha |
Scan area | Cardiac / myocardium |
Scan mode | QuantumPlus |
Acquisition mode | PCCT spectral cardiac acquisition |
Tube voltage [kV] | 120 |
IQ level | 44 CCTA; 55 late iodine enhancement |
Dose modulation | CARE Dose4D |
Rotation time [s] | 0.25 |
Detector slice collimation | 144 × 0.4 mm |
ECG gating | Prospective ECG-triggered |
CCTA ECG window | 30–80% of RR interval |
Late iodine enhancement timing | 5 min after contrast administration |
Late iodine enhancement ECG timing | Fixed 280-ms delay from R-wave |
Baseline slice thickness / increment [mm] | 1.5 / 1.0 |
Baseline kernel | Qr40 |
Baseline iterative reconstruction | QIR 3 |
Baseline reconstruction energy | 65 keV |
QIR levels tested | QIR 1–4 |
Slice thickness / increment tested [mm] | 0.4/0.3; 1.5/1.0; 3.0/2.5; 5.0/4.0; 8.0/7.0 |
VMI energies tested | 40, 45, 50, 55, 60, 65, 70, 90 keV |
Optimal reconstruction | QIR 4; 0.4/0.3 mm; 45 keV |
ECV reconstruction | Iodine maps generated from late iodine enhancement scan |
Contrast agent | Iopromide, Ultravist 370 mg I/mL |
Contrast volume | 100 mL |
Saline flush | 20 mL |
Injection rate | 5–6 mL/s |
Key Results
Parameter | Main finding |
|---|
Study population | 27 patients undergoing same-day PCCT and MRI |
QIR effect | No significant ECV difference across QIR 1–4, p = 1.000 |
Best slice thickness | 0.4 mm, CCC ≈ 0.94, bias 0.8% |
Best VMI energy | 45 keV, mean bias 0.6% |
Optimal reconstruction | QIR 4 + 0.4/0.3 mm + 45 keV |
Improvement vs baseline | 63% lower mean bias and 6% higher concordance with MRI |
Source:
Gnasso C, Pinos D, Schoepf UJ, Vecsey-Nagy M, Aquino GJ, Fink N, Zsarnoczay E, Holtackers RJ, Stock J, Suranyi P, Varga-Szemes A, Emrich T. Impact of reconstruction parameters on the accuracy of myocardial extracellular volume quantification on a first-generation, photon-counting detector CT. European Radiology Experimental. 2024;8:70. doi:10.1186/s41747-024-00469-7.
