Database to CT scan to reduce doses with iterative reconstructions (SAFIRE)

Abstract

Purpose To evaluate the impact of iterative reconstruction SAFIRE on quality image which can be used to reduce dose for CT scan acquisition in clinical practice. Materials and methods 56 CT acquisitions (Somatom Definition AS+, Siemens) were performed on a phantom image quality in several voltages, mAs and pitchs for the same collimation. The raw data were reconstructed by filtered back projection and 5 levels of SAFIRE for two filters and three couples thickness/increment. A total database of 2016 combinations of parameters acquisition and reconstruction was obtained. Noise, signal, contrast to noise ratio (CNR), signal to noise ratio (SNR) and transverse spatial resolution were investigated. From this database, a software tool was developed to find the acquisition and reconstruction parameters for dose reduction while maintaining quality indices equivalent. Concordance of the results was performed on an anthropomorphic phantom for use in clinical practice. Results Reducing the dose by changing the acquisition parameters deteriorated image quality indices, mainly due to the increase in noise. This decline was offset by couples thickness/increment and higher levels SAFIRE. Over the SAFIRE level increased more noise decreased (−10% per level). The signal is stable, the RSB and RCB increased. SAFIRE did not influence the transverse spatial resolution. These results were confirmed with anthropomorphic phantom. The computer tool helping users from the initial settings, can propose to reduce the dose while maintaining adequate RSB and RCB. The best combinations of parameters were proposed. The use of this tool in routine had achieved dose reductions for patients. The thoracic and abdominal CTDIvol acquisitions showed differences respectively −82% and −64% from the values of NRD. Conclusion Multiple combinations of parameters allowed the radiologist through a computer tool, to select a suitable CT scan acquisition with dose reduction in order to obtain adequate quality images.