Analysis of CT acquisition parameters suitable for use in SPECT/CT: A free-response receiver operating characteristic study

PurposeThis dose optimisation free-response receiver operating characteristic (FROC) study compares the diagnostic performance of 8 CT acquisition parameter settings suitable for use with diagnostic quality SPECT/CT systems. Observers were required to localise simulated pulmonary lesions within an anthropomorphic chest phantom.MethodNine observers of varying CT experience (0–10+ years) reviewed 64 cases from each acquisition protocol. FROC methods were used to provide the area under the ROC curve (AUC) figures of merit for each CT acquisition parameter setting. The American College of Radiologists (ACR) CT accreditation phantom was used to acquire quality control (QC) images at each acquisition parameter setting. Measures of contrast to noise ratio (CNR), spatial (contrast) resolution and uniformity could then be compared to the AUC data. Effective dose (E) was estimated using the ImpaCT tool.ResultsAUC analysis revealed no significant difference (p = 0.866) between CT acquisitions for the detection of simulated pulmonary lesions within a chest phantom. The best observer averaged AUC performance was 0.853 (95% CI 0.803–0.903) at an estimated E of 0.82 mSv. The worst observer averaged AUC was recorded at the lowest estimated E (0.28 mSv) was 0.825 (95% CI 0.766–0.885). No correlation was evident between AUC and measures of CNR, spatial resolution or uniformity.ConclusionOur findings suggest that the lowest dose CT acquisition protocol can be used to accurately localise simulated pulmonary lesions within a chest phantom; suggesting that similar protocols could be used on diagnostic quality SPECT/CT systems to allow low-dose evaluations of lung lesions.