Program Information
"Collimation Creep": Evaluation of Technologist Collimation Techniques in a Digital Radiography Suite
B Schwarz*, D Huo , L Rill , University of Florida, Gainesville, FL
Presentations
WE-AB-601-1 (Wednesday, August 2, 2017) 7:30 AM - 9:30 AM Room: 601
Purpose: With the advent of digital radiography (DR), the ability to crop the image post-exposure combined with exposure borders invisible to radiologists has led to less emphasis on collimation, thereby decreasing image quality and increasing patient dose. Without the ability to see exposure borders on images in PACS as in computed radiography (CR), there is reduced quality control on collimation use by technologists. The objective of this study was to investigate collimation differences between post-exposure images and the images viewed by radiologists as well as quantify the number of over- and under-exposures.
Methods: Data from 1,043 exams over a two-month period was exported from an AGFA DXD-500 DR workstation in the pediatric emergency department and analyzed for the following information: anatomy, detector, exposure index (EI), deviation index (DI), auto-collimation vertices (segmented x-ray beam area), and cropped-collimation vertices. The data was parsed from .XML files for each image using MATLAB®. Collimation and exposure data was analyzed for all exams, detector-specific exams, and anatomy-specific exams. Percent differences were calculated between the calculated auto-collimation areas and the cropped-collimation areas selected by the technologist.
Results: For the chest, table, and tethered detectors, the average percent difference between the auto-collimation and cropped-collimation were -9.8%, -5.8%, and -13.9%, respectively, with maximum percent differences of -70.8%, -77.9%, and -62.6%, respectively. The percentage of exams where the percent difference between the auto-collimation and cropped-collimation was significant (≥5%) was 33.9%, 24.1%, and 44% for the chest, table, and tethered detectors, respectively. Additionally, 22.3% of all exams resulted in an over-exposure (DI≥3) and 3.8% of all exams resulted in an under-exposure (DI≤-3). Exposure-related data was similarly generated for detector-specific exams and anatomy-specific exams.
Conclusion: Application of this data has the potential to improve technologist technique, improve image quality, and reduce patient dose owing to more precise collimation during pediatric exposures.
Contact Email: