Assessment of an iterative reconstruction algorithm (SAFIRE) on image quality in pediatric cardiac CT datasets

Abstract

Background: Pediatric cardiac patients often undergo repeat diagnostic testing, resulting in relatively high cumulative medical radiation exposure. Low-dose CT scanning techniques used to decrease radiation exposure may result in reduced image quality.

Objective: This study evaluates a prototype iterative reconstruction algorithm, sinogram-affirmed iterative reconstruction (SAFIRE), to determine the effect on qualitative and quantitative measures of image quality in pediatric cardiac CT datasets, compared with a standard weighted filtered back projection (wFBP) algorithm.

Methods: Seventy-four datasets obtained on a 128-slice dual-source CT system were evaluated for image quality using both the wFBP and the prototype iterative reconstruction algorithm. Contrast, noise, contrast-to-noise ratio, signal-to-noise ratio, and qualitative image quality were compared between groups. Data were analyzed as medians and 25th and 75th percentiles, and groups were compared with the use of the Wilcoxon singed-rank test or k sample equality of medians test.

Results: There was a 34% decrease in noise, a 41% increase in contrast-to-noise ratio, and a 56% increase in signal-to-noise ratio in the prototype iterative reconstruction, compared with wFBP. All differences were statistically significant (P < 0.001). Qualitative measures of image noise and noise texture were also improved in the iterative reconstruction group (P < 0.001 for both). Diagnostic confidence was similar between reconstruction techniques. Median scan dose length product was 15.5 mGy · cm.

Conclusion: The prototype iterative reconstruction algorithm studied significantly reduces image noise and improves qualitative and quantitative measures of image quality in low-dose pediatric CT datasets, compared with standard wFBP.