Performance of virtual non-contrast images generated on clinical photon-counting detector CT for emphysema...

To evaluate the performance of virtual non-contrast images (VNC) compared to true non-contrast (TNC) images in photon-counting detector computed tomography (PCD-CT) for the evaluation of lung parenchyma and emphysema quantification.

65 (mean age 73 years; 48 male) consecutive patients who underwent a three-phase (non-contrast, arterial and venous) chest/abdomen CT on a first-generation dual-source PCD-CT were retrospectively included. Scans were performed in the multienergy (QuantumPlus) mode at 120 kV with 70 ml intravenous contrast agent at an injection rate of 4 ml s^-1. VNC were reconstructed from the arterial (VNC_art) and venous phase (VNC_ven). TNC and VNC images of the lung were assessed quantitatively by calculating the global noise index (GNI) and qualitatively by two independent, blinded readers (overall image quality and emphysema assessment). Emphysema quantification was performed using a commercially available software tool at a threshold of -950 HU for all data sets. TNC images served as reference standard for emphysema quantification. Low attenuation values (LAV) were compared in a Bland-Altman plot.

GNI was similar in VNC_art (103.0 ± 30.1) and VNC_ven (98.2 ± 22.2) as compared to TNC (100.9 ± 19.0, p = 0.546 and p = 0.272, respectively). Subjective image quality (emphysema assessment and overall image quality) was highest for TNC (p = 0.001), followed by VNC_ven and VNC_art. Both, VNC_art and VNC_ven showed no significant difference in emphysema quantification as compared to TNC (p = 0.409 vs. p = 0.093; respectively).

Emphysema evaluation is feasible using virtual non-contrast images from PCD-CT.

Emphysema quantification is feasible and accurate using VNC images in PCD-CT. Based on these findings, additional TNC scans for emphysema quantification could be omitted in the future.