Establishment of Submillisievert Abdominal CT Protocols With an In Vivo Swine Model and an Anthropomorphic Phantom.

OBJECTIVE. The purpose of this study was to evaluate radiation dose, effective dose, and image quality of different low-dose abdominal CT protocols in a swine model and an anthropomorphic phantom using a third-generation dual-source CT scanner. MATERIALS AND METHODS. Four different abdominal low-dose protocols were established using a swine model and were regarded as diagnostic by two experienced radiologists on the basis of clarity and sharpness of anatomic structures. General image conditions such as noise and spatial resolution as well as diagnostic acceptability and artifacts were evaluated. Objective image quality was determined by measuring signal-to-noise ratio (SNR) and contrast-to-noise ratio (CNR) in different anatomic locations. To evaluate the effective dose, thermoluminescent dosimeter (TLD) measurements were repeated in a phantom. RESULTS. Diagnostic acceptability, spatial resolution, and noise were rated as optimal in all four protocols, which were therefore regarded as diagnostic. We found no statistically significant differences in SNR or CNR for the four low-dose protocols. Effective dose determined from the phantom measurements did not exceed 0.33 mSv for any protocol. Overall evaluation of the 86 TLD measurements for the four low-dose protocols revealed a statistically significant difference in radiation dose (p < 0.0001), showing that the dual-source protocol had the lowest radiation dose. CONCLUSION. Submillisievert abdominal CT is feasible with good image quality and doses even lower than conventional abdominal radiography. Our dual-source protocol achieved the lowest dose, which further shows that dual-source imaging is possible in the sub-millisievert range without additional dose.