Abstracts

RETURN TO ABSTRACT LISTING


1844. Optimizing Identification of Power Injectable Ports on the Scout Images for MDCT Procedures
Authors * Denotes Presenting Author
  1. Fides Schwartz *; Duke University Health System
  2. Kevin Iliffe; Duke University Health System
  3. Andrew Somers; Duke University Health System
  4. Brittney Stone; Duke University Health System
  5. Tracy Jaffe; Duke University Health System
  6. Justin Solomon; Duke University Health System
  7. Lynne Hurwitz Koweek; Duke University Health System
Objective:
Assess the impact of tube voltage and image display on the identification of power ports features on AP scout images to inform optimal workflow for MDCT exams.

Materials and Methods:
Four ports, representing variable material composition (titanium/silicone), shapes and CT markings, were imaged on an adult anthropomorphic chest phantom (LUNGMAN 1, PH-1 41337-000, Kyoto Kagaku Co., Ltd., Kyoto, Japan) using a dual-source MDCT at variable peak tube voltages (80, 100, 120, 150, Sn150 kVp). Fourty-six images were reviewed at variable image display setting by 5 blinded readers (two CT technologists (9 and 34 years of experience), one radiology nurse (20 years of experience), and two fellowship trained radiologists (23 and 24 years of experience)) to assess port features of material composition, shape and text markings as well as overall preferred image quality. Descriptive statistics were applied, inter-reader agreement was analyzed using Cohen's kappa (1-2) and magnitude of agreement was evaluated based on Landis and Koch (3).

Results:
Material composition was correctly identified for all ports by all readers across all kVp settings. The identification by shape was more reliable than CT markers for all but one of the ports. CT marker identification was up to 80% for titanium ports at window level settings optimized for metal (W: 200, C: -150) and at a soft tissue setting (W: 400, C: 40) for silicone ports. Inter-reader agreement for best image quality per kVp setting was moderate to substantial for three ports (k=0.5-0.62) but only fair for one port (k=0.27). The highest overall rank for image quality was given unanimously to Sn150 kVp for imaging titanium ports and 100 kVp for silicone ports.

Conclusion:
Power port identification on MDCT scout images can be optimized with modification of MDCT scout acquisition and display settings based on the main port material. Appropriate identification of ports for infusion of contrast at high injection rates using the scout images can improve workflow, avoiding delays by sending the patient for an x-ray prior to the MDCT scan or gaining alternate access.