Optimum CT reconstruction parameters for vascular and hepatocellular carcinoma models in a liver phantom with multi-level dynamic computed tomography with 64 detector rows: a basic study.

We quantified to clarify the optimum factors for CT image reconstruction of an enhanced hepatocellular carcinoma (HCC) model in a liver phantom obtained by multi-level dynamic computed tomography (M-LDCT) with 64 detector rows. After M-LDCT scanning of a water phantom and an enhanced HCC model, we compared the standard deviation (SD, 1 ± SD), noise power spectrum (NPS) values, contrast-noise ratios (CNR), and the M-LDCT image among the reconstruction parameters, including the convolution kernel (FC11, FC13, and FC15), post-processing quantum filters (2D-Q00, 2D-Q01, and 2D-Q02) and slice thicknesses/slice intervals. The SD and NPS values were lowest with FC11 and 2D-Q02. The CNR values were highest with 2D-Q02. The M-LDCT image quality was highest with FC11 and 2D-Q02, and with slice thicknesses/slice intervals of 0.5 mm/0.5 mm and 0.5 mm/0.25 mm. The optimum factors were the FC11 convolution kernel, 2D-Q02 quantum filter, and 0.5 mm slice thickness/0.5 mm slice interval or less.