Reducing the radiation dose with the adaptive statistical iterative reconstruction technique for chest CT in adults: a parameter study.

Background: Currently there is a trend towards reducing radiation dose while maintaining image quality during computer tomography (CT) examination. This results from the concerns about radiation exposure from CT and the potential increase in the incidence of radiation induced carcinogenesis. This study aimed to investigate the lowest radiation dose for maintaining good image quality in adult chest scanning using GE CT equipment.

Methods: Seventy-two adult patients were examined by Gemstone Spectral CT. They were randomly divided into six groups. We set up a different value of noise index (NI) when evaluating each group every other number from 13.0 to 23.0. The original images were acquired with a slice of 5 mm thickness. For each group, several image series were reconstructed using different levels of adaptive statistical iterative reconstruction (ASIR) (30%, 50%, and 70%). We got a total of 18 image sequences of different combinations of NI and ASIR percentage. On one hand, quantitative indicators, such as CT value and standard deviation (SD), were assessed at the region of interest. The signal-to-noise ratio (SNR) and contrast-to-noise ratio (CNR) were calculated. The volume CT dose index (CTDI) and dose length product (DLP) were recorded. On the other hand, two radiologists with ≥ 5 years of experience blindly reviewed the subjective image quality using the standards we had previously set.

Results: The different combinations of noise index and ASIR were assessed. There was no significant difference in CT values among the 18 image sequences. The SD value was reduced with the noise index's reduction or ASIR's increase. There was a trend towards gradually lower SNR and CNR with an NI increase. The CTDI and DLP were diminishing as the NI increased. The scores from subjective image quality evaluation were reduced in all groups as the ASIR increased.

Conclusions: Increasing NI can reduce radiation dose. With the premise of maintaining the same image quality, using a suitable percentage of ASIR can increase the value of NI. To assure image quality, we concluded that when the NI was set at 17.0 and ASIR was 50%, the image quality could be optimal for not only satisfying the requirements of clinical diagnosis, but also achieving the purpose of low-dose scanning.