Data acquisition and reconstruction
All imaging data were acquired using a Symbia Intevo16 hybrid SPECT/CT system (Siemens Healthineers, Erlangen, Germany) comprising an integrated dual-head SPECT camera with a 16-slice helical CT scanner. We acquired SPECT images under the following parameters: ±7.5% energy window at 140 keV with a lower scatter window of 15%, ⅜” crystal thickness, low-energy high-resolution collimator, 256 × 256 matrix with 2.4-mm pixels and a total of 120 projections of 15 s/view over 360° in a non-circular orbit continuous acquisition mode. Immediately following the SPECT acquisition, CT images were acquired at 130 KV and 70 ref mA using adaptive dose modulation (CARE Dose 4D; Siemens Healthineers) with a 512 × 512 matrix, pitch 1.5, 0.8-s rotation and 2 × 1.5-mm collimation. The CT data were reconstructed at a 3.0-mm slice thickness using a B31s attenuation filter (Siemens Healthineers).
We reconstructed the SPECT images using the algorithms F3D, xQ and xB and a 6-mm 3D Gaussian filter with various combinations of one fixed subset and 1-96 iterations. The F3D is equipped with OSEM and depth-dependent 3D resolution recovery using the Gaussian point-spread functions, AC and SC. The xQ and xB are equipped with OSCGM and depth-dependent 3D resolution recovery using AC and SC. The xB algorithm divides CT pixels into six tissue classes with smooth boundaries based on CT values or “zones” of air and lung, adipose, soft tissue, soft bone, cortical bone, metal material, and updates. The xB iterative operation can be weighted according to the corresponding zone class in the divided pixel, however, the iterative operation for each zone class based on the CT data showed no effect in raising the original count [24].
Cross-calibration of SPECT imaging
Reconstructed SPECT counts derived from F3D and xSPECT were converted to activity concentrations based on a cross-calibration factor (CCF) obtained from the relationship between the reconstructed counts and activity concentrations and system planar sensitivity, respectively, for quantitative comparisons.
In SPECT images using F3D, a circular region of interest (ROI) to measure SPECT count density (counts/mL) was placed at the center of the cylindrical phantom on the central slice and at ±1 and ±2 slices from the center. The CCF was automatically calculated using GI-BONE software (Aze, Tokyo, Japan) software as the ratio of the actual activity concentration (measured by the dose calibrator) in the phantom at the time of scanning to the measured SPECT counts density per scan duration [30]. The using dose calibrator used for cross-calibration was CRC-15R (final calibration date by manufacturer: 4/19/2005). The dose calibrator is also confirmed and calibrated with a cite-specific NIST traceable 68Ge/68Ga source every 3 months (final calibration date in site: 12/18/2019) [31, 32]. Therefore, we consider that the uncertainty of the measurement by the dose calibrator is small. The actual SUV was calculated as:
Reconstruction with xQ and xB precisely determine images in units of Bq/mL that are converted using system planar sensitivity with NIST traceable 57Co source [21]. The system planar sensitivity is a necessary parameter to allow for the conversion between the count rate and units of absolute activity. This is defined as a measure of how many counts the gamma camera detects for every unit of activity in its field of view. Therefore, system planar sensitivity was measured with traceable point source without scattering and attenuation to realize accurate and reproducible quantitation [28, 33]. The use of this source is recommended for all Siemens® users to improve SPECT quantification. We automatically converted the quantitative SPECT/CT data using MI Applications VB10 (Siemens Healthineers).
Phantom Studies
Phantom design
We custom-designed a physical three-dimensional phantom to determine the bone SPECT-specific distribution of activity and the linear attenuation coefficient (Figure 1). This phantom can be used to generate SPECT images of bone metastasis with a realistic abdomen contour [34]. The phantom contains a 99mTc solution to simulate the soft tissue, and the vertebral body, spinous and transverse process and tumor region contained a bone-equivalent solution of K2HPO4 and 99mTc [35]. The phantom experiments proceeded twice using different activity concentrations and sphere sizes as follows. Tumor, normal bone and soft tissue in the phantom are filled with 99mTc solution. In the first round of experiments, a body phantom with four 28-mm diameter spheres was set and acquired at tumor-to-normal bone ratios (TBR) of 1, 2, 4 and 10 at a normal bone activity level of 50 kBq/mL. This phantom contained 8 kBq/mL of a 99mTc solution as the background activity of the soft tissue. That is, in the case of TBR1, there is no difference between the boundary and the background, and the difference in the activity concentration increases as a function of the higher TBR. In the second round of experiments, another phantom was set as 13-, 17-, 22-, 28-mm diameter spheres and used to determine the activity concentrations of the simulated soft tissue, normal bone and tumor at 8, 50 and 200 kBq/mL, respectively, i.e., TBR4.
Figure 1
Data Analysis
The SPECT acquisition data in the first round of experiments were reconstructed using subset 1 and 1-96 iterations. We examined the effects of the reconstruction algorithms on various TBR in the 28-mm sphere and then determined the optimal reconstruction parameters in accordance with the result of this convergence characteristic. Phantom images containing different sizes of simulated tumors were continuously analyzed in terms of the spatial resolution of a 10-mm spinous process, the coefficient of variance (CV) and the contrast-to-noise ratio (CNR) of the vertebral body and RC as quantitative parameters. We drew profile curves on the spinous process, measured the full width at half maximum (FWHM), and evaluated the CV at an 80% circular ROI (ROI80%) placed at the center of the vertebral body. The RC were placed at circular ROI with diameters of 13, 17, 22, 28 mm. The CV was calculated as SD/mean, where SD represents the standard deviation of the ROI in the radioactive section and mean represents the mean SPECT value (kBq/mL) in the ROI. The CNR at TBR 4 was calculated as (Hs-Hnb) / σnb, where, Hs and Hnb are the activity concentrations measured in the spheres and normal bone, respectively, and σnb is the voxel SD in the normal bone. The RC was defined as the ratio of the mean SPECT value (kBq/mL) and the true SPECT value (kBq/mL) for each sphere.
Clinical study
Imaging protocol
We analyzed data from 20 consecutive patients who had undergone bone SPECT/CT imaging for metastatic prostate or breast cancer (male, n = 13; female, n = 7; median age, 62 years; range, 40–83 years; average weight, 65.2 ± 13.4 kg; range, 51.8–78.6 kg). The optimal condition of the convergence characteristic in the phantom study was applied in the clinical reconstruction parameters in xQ and xB. Bone SPECT/CT imaging proceeded from the abdomen to the pelvis ~2.5–4 h after delivering an intravenous injection of 1003.4 ± 102.8 MBq 99mTc-methylene diphosphonate (99mTc-MDP; FUJIFILM Toyama Chemical, Tokyo, Japan) or hydroxymethylene diphosphonate (99mTc-HMDP; Nihon Medi-Physics, Tokyo, Japan). The average amount of injected 99mTc was 15.9 ± 2.8 (range, 13.1–18.7) MBq/kg. The Ethics Committee at the Cancer Institute Hospital of JFCR approved this clinical study (approval no. 2015-1151). These clinical data were analyzed in this retrospective study and the results did not influence any further therapeutic decision-making.
Data analysis
The noise characteristics and quantitative performance of the clinical SPECT image were analyzed at the level of the 4th vertebral body (L4) [36]. We placed a ROI80% on the center of the axial slice in the vertebral body section, and a ROI80% exactly on the corresponding vertebral body in the central slice by following the CT boundaries of the fused SPECT/CT images (Figure 2). We determined as the SUVmax, SUVmean and SUVpeak as normalized by the patient’s body weight. These data were analyzed using PETSTAT software (AdIn Research, Tokyo, Japan).
Figure 2
Statistical analysis
All SUV and CV indices in the xQ and xB groups were compared using Wilcoxon signed-ranks tests after evaluating the non-normal distribution using Kolmogorov-Smirnov tests. Values were considered statistically significant when P < 0.05. These data were statistically analyzed using SPSS Statistics software (IBM, Armonk, NY, USA).