Comparison of filtered back projection and iterative reconstruction in diagnosing appendicitis at 2-mSv CT

Abstract

Purpose: To compare radiologists' diagnostic performance and confidence, and subjective image quality between filtered back projection (FBP) and iterative reconstruction (IR) at 2-mSv appendiceal CT.

Methods: The institutional review board approved this retrospective study and waived the requirement for informed consent. We included 107 adolescents and young adults (age, 29.8 ± 8.5 years; 64 females) undergoing 2-mSv CT for suspected appendicitis. Appendicitis was pathologically confirmed in 42 patients. Seven readers with different experience levels independently reviewed the CT images reconstructed using FBP and IR (iDose(4), Philips). They rated both the likelihood of appendicitis and subjective image quality on 5-point Likert scales. Diagnostic confidence was assessed using the likelihood of appendicitis, proportion of indeterminate interpretations, and 3-point normal appendix visualization score. We used receiver operating characteristic analyses, Wilcoxon's signed-rank tests, and McNemar's tests.

Results: The pooled area under the receiver operating characteristic curve (AUC) was 0.96 for both FBP and IR (95% CI for the difference, -0.02, 0.02; P = 0.73). The AUC difference was not significant in any of the individual readers (P ≥ 0.21). For the majority of the readers, the diagnostic confidence was not significantly different between the two reconstruction methods. Subjective image quality tended to be higher with IR for all readers (P ≤ 0.70), showing significant differences for four readers (P ≤ 0.040).

Conclusion: When diagnosing appendicitis at 2-mSv CT in adolescents and young adults, FBP and IR were comparable in radiologists' diagnostic performance and confidence while IR exhibited higher subjective image quality than FBP.

Keywords: Appendicitis; Computer assisted; Image processing; Iterative reconstruction; ROC curve; Tomography; X-ray computed.