Quantitative CT Detectability of PEGylated Gold Nanoparticles: Influence of Reconstruction Parameters

Gold nanoparticles (AuNPs) have attracted considerable interest as emerging CT contrast agents; however, the influence of reconstruction parameters on their quantitative detectability remains insufficiently characterized, limiting the optimization of quantitative CT imaging. This study quantitatively compared the CT imaging performance of citrate-stabilized AuNPs, PEGylated AuNPs (PEG-AuNPs), and a conventional iodinated contrast agent (BMX) under different reconstruction conditions. AuNPs were synthesized by citrate reduction and PEGylated with SH-PEG5K-COOH, followed by physicochemical characterization using dynamic light scattering, zeta potential analysis, and ICP-OES. Contrast samples (≤10 mg/mL) were imaged using a micro-CT system (50 kV, 0.43 mA) with two angular sampling intervals (0.750° and 0.500°), Gaussian and Hann reconstruction kernels, and voxel sizes of 100, 160, and 200 µm. Quantitative performance was evaluated using sensitivity, coefficient of determination (R2), background noise, limit of detection (LOD), and limit of quantification (LOQ). PEG-AuNPs exhibited the highest sensitivity (80.55 HU/mg/mL) and the lowest detection limits (LOD: 0.53 mg/mL; LOQ: 1.61 mg/mL), outperforming bare AuNPs and BMX. Voxel size was the dominant reconstruction parameter, with optimal detectability achieved at 160 µm due to an optimal balance between spatial resolution and image noise, whereas reconstruction kernel and angular sampling had comparatively minor effects. These findings provide a quantitative framework for optimizing reconstruction parameters in nanoparticle-enhanced quantitative CT imaging.