ARRS 2022 Abstracts


E1981. Image Quality Assessment of Pediatric Abdominal CTA
  1. Shyam Sunder Venkatakrishna; The Children's Hospital of Philadelphia
  2. Abhay Srinivasan; The Children's Hospital of Philadelphia; University of Pennsylvania, Perelman School of Medicine
  3. Monica Miranda-Schaeubinger; The Children's Hospital of Philadelphia
  4. Michael Acord; The Children's Hospital of Philadelphia; University of Pennsylvania, Perelman School of Medicine
  5. Patricia Mecca; The Children's Hospital of Philadelphia
  6. Xiaowei Zhu; The Children's Hospital of Philadelphia
  7. Summer Kaplan; The Children's Hospital of Philadelphia; University of Pennsylvania, Perelman School of Medicine
Abdominal CTA is often critical for the assessment of pediatric vasculature, especially for liver transplant planning. Pediatric CTA protocols have often emphasized reducing radiation dose, but low-dose techniques can compromise image quality. We assessed differences between subjective image quality and objective measures of contrast-to-noise ratio (CNR), signal-to-noise ratio (SNR), and pediatric size-specific dose estimate (SSDE).

Materials and Methods:
A retrospective review of pediatric abdominal CTA performed between 2018 to 2020 was conducted. Two pediatric radiologists assessed image quality as acceptable or unacceptable on arterial, portal, and venous phases; a study was deemed unacceptable if at least one radiologist noted unacceptability. CNR and SNR were calculated on a reference index slice using liver vasculature as the target and paraspinal muscle as background. SSDE and patient demographics were recorded. We compared age, CNR, SNR, and SSDE between acceptable and unacceptable groups.

A total of 69 patients underwent abdominal CTA, with age range of 0.08–19 years, median age 2 years. Mottle was acceptable in 53/69 (77%) arterial phases, 35/65 (54%) portal phases, and 23/59 (39%) venous phases. There was a significant difference in median ages of patients with regard to subjective image quality acceptability in arterial and portal phases (1.8 y for acceptable vs. 12.8 y for unacceptable). Similarly, SNR showed a significant difference between groups for all phases (all p < 0.005). CNR was different between acceptability groups for portal phase only (p < 0.001). A significant difference in SSDE between acceptability groups was not demonstrated for hepatic (1.8 vs 2.1 mGy) or portal venous phases (1.8 vs. 2 mGy). Results suggest that SNR, a measure of quantum mottle, may be an objective marker of subjectively acceptable image quality for pediatric abdominal CTA. CNR and SSDE were not consistently different between acceptable and unacceptable groups for all phases, suggesting that contrast enhancement and dose are less reliable objective indicators of image quality. Because patients with unacceptable image quality were older, the similar SSDE between the groups may indicate underexposure for older, larger patients.

Determining a SNR threshold for acceptable image quality may help radiology staff maintain image quality without the need for radiologist subjective image review, while keeping CTA dose within appropriate parameters.