Determining the calibration factor of a SPECT/CT camera

Aim Quantitative single-photon emission computed tomography (SPECT) has several important applications including monitoring tumor response after treatment and dose estimation for targeted radionuclide therapy treatment planning. The use of a calibration factor is required to obtain quantitative SPECT images. In this study, the calibration factor (CF) was determined for different acquisition and reconstruction protocols. Materials and methods A cylindrical phantom (height: 19.5 cm , diameter: 19.5 cm) was used for the SPECT/CT acquisitions (GE Discovery NM/CT 670). The phantom was filled with 99mTc, approximately 320 MBq for the first measurements and approximately 230 MBq for the second measurement that was performed to check the reproducibility of the calibration factor. The phantom was acquired four times using different acquisition parameters (change of acquisition time, matrix size, zoom factor) and the phantom was reconstructed using the GE Xeleris Volumetrix and GE Xeleris Volumetrix Evolution for Bone software. A volume of interest (VOI) (height: 15 cm, diameter: 17 cm) was drawn on the acquired images. Knowing the counts in the VOI, the volume of the VOI and the acquisition time, the count rate per unit volume was calculated. The CF was then determined by dividing the count rate per unit volume by the activity concentration decay corrected at the beginning of the acquisition. Results The phantom reconstructed with the GE Xeleris Volumetrix resulted in a mean CF of 7.93 cpm/kBq (counts per minute per kilo Bequerel) for the 4 different acquisition protocols (CV: 0.04). Repeating this experiment resulted in a mean CF of 7.98 cpm/kBq (CV: 0.06) which only deviates 0.61% to the CF obtained during the first experiment. The phantom reconstructed with the GE Xeleris Volumetrix Evolution for Bone resulted in a mean CF of 40.51 cpm/kBq (counts per minute per kilo Bequerel) for the 4 different acquisition protocols (CV: 0.02). Repeating this experiment resulted in a mean CF of 41.29 cpm/kBq (CV: 0.02) which only deviates 1.93% to the CF obtained during the first experiment. Conclusion The CF is not influenced by acquisition time, matrix size or zoom factor, but is highly influenced by the reconstruction parameters, more specifically resolution recovery (used in the GE Xeleris Volumetrix Evolution for Bone software). Therefore it is highly important when using a CF clinically to reconstruct the phantom data with the same reconstruction parameters as patient data.status: publishe