DECT acquisition and reconstruction
A 32-cm-diameter water phantom was scanned using second-generation dual-layer detector computed tomography at 100, 120, and 140 kVp. All images were acquired using an 80-mm-wide helical scanning mode with 128 × 0.625 mm collimation. The tube current was adjusted to achieve volume computed tomography dose indices (CTDIvol) of 10 and 20 mGy at each tube voltage. Both small and large focal-spot sizes were employed for each CTDIvol level. Ten scans were performed to measure the magnitude of noise. The additional detailed parameters are listed in Table 1. VMIs at 70 keV were reconstructed for each tube voltage.
Measurement of noise magnitude
Noise magnitude was assessed by calculating the standard deviation (SD) within five regions of interest (ROIs), each comprising approximately 1% of the phantom's cross-sectional area. One ROI was placed at the center, whereas the remaining ROIs were placed at the 12, 3, 6, and 9 o’clock positions. The peripheral ROIs were placed approximately one ROI diameter from the phantom border (Figure 1) [10]. The SD of pixel values was recorded at each ROI and then averaged across 10 images; this average pixel SD served as the noise magnitude output. The SDs of the 70 keV VMIs reconstructed at each tube voltage were compared among CTDIvol levels with respect to focal-spot size. To evaluate the effect of focal spot size on SD, the SD ratio was calculated by dividing the SD obtained with the small focal spot by the SD obtained with the large focal spot at the same CTDIvol.