Michael James Callahan MD, Presenter: Nothing to Disclose

Patricia Kleinman MPH, Abstract Co-Author: Nothing to Disclose

Keith J. Strauss MS, Abstract Co-Author: Nothing to Disclose

Andriy I. Bandos PhD, Abstract Co-Author: Nothing to Disclose

Paul K. Kleinman MD, Abstract Co-Author: Nothing to Disclose

Andy Tsai MD, Abstract Co-Author: Nothing to Disclose

George A. Taylor MD, Abstract Co-Author: Nothing to Disclose

Determine effect of simulated CT dose reduction on detection and localization of the normal appendix in cases of suspected appendicitis in children.

201 consecutive pediatric dose modulated CT examinations performed for suspected appendicitis were reviewed. Based on clinical follow-up, 45 cases were selected where the diagnosis of appendicitis was excluded. Upon review of all axial (0.625mm, 2.5mm, 5.0mm), coronal (3.0mm) and sagittal (3.0mm) reformatted images, 2 pediatric radiologists (judges) determined visualization of the appendix in 31 cases and non-visualization in 14. Judges placed an arrow on the appendix and cross referenced appropriate images. For each of the 45 cases, noise was introduced to the 5.0mm axial images (100% dose) by a computer-simulated dose reduction software program to generate 5 additional axial series ranging from 76% to 25% of the original mAs. 270 axial 5.0mm CT series (45 x 6) were randomized. Using a 4-point Likert scale and arrow localization, the original and dose reduced images were reviewed by 2 pediatric radiologists with 2 and 25 years experience.

Decreasing confidence ratings were noted for decreasing exposure levels, which were statistically significant for both appendix-visualized (average difference varying from -.39 to .11 points on 0-3 scale, p=0.026) and non-visualized appendix cases (-.24 to 0.11, p=0.035). Average ability of raters to discriminate between appendix-visualized and non-visualized cases did not exhibit a significant change under different exposure levels. The area under the ROC curve (AUC) ranged from 0.74 to 0.79, with 0.76 for the full dose. There was no observed decrease in AUCs for exposure levels < 50%, with the 95% confidence intervals for the difference residing in a -0.09 to 0.14 range. There were no significant differences between the full-dose and dose–reduced series in terms of the numbers of the correctly localized appendices or the number of erroneously marked non-visualized appendix cases.

Standard pediatric CT scan mAs can be substantially reduced without significantly affecting a radiologist's performance in the identification of a normal appendix in cases of suspected appendicitis.

Appendicitis is the most common surgical emergency in children. A judicious reduction in CT dose for suspected appendicitis may have a positive effect on cancer risks in the pediatric population.

Callahan, M, Kleinman, P, Strauss, K, Bandos, A, Kleinman, P, Tsai, A, Taylor, G, Simulated CT Dose Reduction for Detection of the Normal Appendix in Cases of Suspected Appendicitis in Children.  Radiological Society of North America 2011 Scientific Assembly and Annual Meeting, November 26 - December 2, 2011 ,Chicago IL. http://archive.rsna.org/2011/11015680.html