Validation of Computed Tomography-based Attenuation Correction of Deviation between Theoretical and Actual Values in Four Computed Tomography Scanners

OBJECTIVE(S): In this study, we aimed to validate the accuracy of computed tomography-based attenuation correction (CTAC), using the bilinear scaling method. METHODS: The measured attenuation coefficient (μ(m)) was compared to the theoretical attenuation coefficient (μ(t)), using four different CT scanners and an RMI 467 phantom. The effective energy of CT beam X-rays was calculated, using the aluminum half-value layer method and was used in conjunction with an attenuation coefficient map to convert the CT numbers to μ(m) values for the photon energy of 140 keV. We measured the CT number of RMI 467 phantom for each of the four scanners and compared the μ(m) and μ(t) values for the effective energies of CT beam X-rays, effective atomic numbers, and physical densities. RESULTS: The μ(m) values for CT beam X-rays with low effective energies decreased in high construction elements, compared with CT beam X-rays of high effective energies. As the physical density increased, the μ(m) values elevated linearly. Compared with other scanners, the μ(m) values obtained from the scanner with CT beam X-rays of maximal effective energy increased once the effective atomic number exceeded 10.00. The μ(m) value of soft tissue was equivalent to the μ(t) value. However, the ratios of maximal difference between μ(m) and μ(t) values were 25.4% (lung tissue) and 21.5% (bone tissue), respectively. Additionally, the maximal difference in μ(m) values was 6.0% in the bone tissue for each scanner. CONCLUSION: The bilinear scaling method could accurately convert CT numbers to μ values in soft tissues.