Purpose or Case Report: To evaluate the effect of different levels of adaptive statistical iterative reconstruction (ASIR) in pediatric lower extremity CT scan to compare the radiation dose, objective imaging parameters and subjective diagnostic image quality ratings.
Methods & Materials: This HIPPA-compliant, quality improvement study was approved for IRB-exempt status. The study cohort included 91 pediatric patients who underwent CT scan of the lower extremity, defined as a CT of the thigh, knee, leg, ankle or foot. The control group consisted of 37 patients who were scanned with a standard CT protocol using 0% ASIR and 100% FBP. The other group of patients was imaged on a CT scanner equipped with ASIR software. Out of the second group 20 patients were scanned with 30% ASIR and 70% FBP, and 34 patients were scanned with 40% ASIR and 60% FBP.
Objective imaging parameters obtained on all studies included signal-to-noise ratio (SNR) within bone versus air and contrast-to-noise ratio (CNR) between bone and muscle versus air. All measurements were made using standardized regions of interest (1-1.5cm²) on images of 1.25mm slice thickness. Computed tomography dose index (CTDI) and dose-length product (DLP) were recorded for each study as estimates of radiation dose.
Two pediatric radiologists who were blinded to the level of ASIR made subjective ratings on each study with agreement. The categories of rating included image noise, image sharpness, diagnostic acceptability and artifacts.
The objective data was analyzed by linear regression and the subjective data from the readers was analyzed by Logistic regression that compared the level of ASIR to the image quality ratings of each category.
Results: We found significant radiation dose reduction with higher levels of ASIR, indicated by decrease in CTDI (p-value 0.0046) and DLP (p<0.001). With increased ASIR levels, subjective noise level decreased (p<0.0001) and signal-to-noise ratio increased (p-value 0.0006). As the ASIR level increased, however, the image sharpness and diagnostic acceptability decreased on subjective assessment by readers (p< 0.0001 and p-value 0.0100 respectively) compared to standard CT protocols studies.
Conclusions: Pediatric lower extremity CT performed at 30% and 40% ASIR levels significantly reduced the radiation dose and perceived noise level with improved signal to noise ratio (SNR) compared to standard CT scan. The subjective image sharpness and diagnostic acceptability, however, significantly decreased as the ASIR level increased.