Comparison of standard-dose and half-dose dual-source abdominopelvic CT scans for evaluation of acute abdominal pain

Abstract

Background With the increasing number of computed tomography (CT) scans used for evaluation of acute abdominal pain, patient radiation exposure has increased rapidly. Purpose To determine whether the diagnostic performance of half-dose abdominopelvic CT is non-inferior to that of standard-dose CT for patients with acute abdominal pain. Material and Methods Ninety-eight patients with acute abdominal pain underwent dual-source abdominopelvic CT. Three sets of CT images were reconstructed: standard-dose filtered back projection (FBP); half-dose FBP; and half-dose sinogram-affirmed iterative reconstruction (SAFIRE(3)). Diagnostic performance of the standard-dose scan was compared with that of the half-dose scans by using a non-inferiority test with a 10% margin. The overall image quality was subjectively measured. Results Diagnostic performance for overall disease diagnosis with half-dose scans (area under the receiver operating characteristic curve [AUC] = 0.835 for FBP, 0.881 for SAFIRE(3)) was non-inferior to that of standard-dose FBP (AUC = 0.891) (95% confidence interval lower limit difference = -5.6% [half-dose FBP], -1.2% [half-dose SAFIRE(3)]). The diagnostic sensitivity for detection of neoplastic disease was lower with half-dose (75.0%) than with standard-dose FBP (91.7%). Effective dose and dose-length product with standard-dose imaging were 7.99 +/- 2.55 mSv and 533.1 +/- 170.3 mGy center dot cm, respectively; those of half-dose imaging were 3.99 +/- 1.28 mSv and 266.6 +/- 85.2 mGy center dot cm, respectively. The image quality was lower with half-dose than with standard-dose FBP scans (P < 0.01). Conclusion Diagnostic performance of half-dose CT is non-inferior to that of standard-dose scan for evaluation of acute abdominal pain, despite inferior image quality.