Purpose: In digital radiography (DR), ongoing analysis of exposure index (EI) and deviation index (DI) serves to identify incidences of over- and under-exposure. It further allows DI control limits and EI targets (EIT) to be created/improved. The goal of this study was to conduct an initial analysis for one type of DR system to serve as a performance baseline for ongoing quality improvement.
Methods: The study included exposure data from 32,012 adult radiographic exams acquired using three DR systems (DiscoveryXR656HD or OptimaXR646HD, GE Healthcare). For each anatomical view with at least 1000 data points, we characterized the central tendency and the dispersion of EI, EIT, and DI. Percentage of exams with DI outside of the manufacturer-defined acceptable range of [-5,4] was determined. If the mean DI approached zero (within ±1), initial DI control limits were established following AAPM TG232 recommendation. Otherwise, the need to improve practice or revise EIT was investigated.
Results: Nine anatomical views had 1000 data points or more. Median EI values ranged between 139 for posteroanterior (PA) chest to 376 for anteroposterior (AP) upper rib with interquartile range of 55 and 243, respectively. For a given view, EIT was not constant. For example, the EIT for AP lumber spine ranged between 269 and 405, increasing approximately linearly with kVp. Percentage of exams with DI outside of the manufacturer-defined acceptable range varied from 0.6% for PA chest to 73% for lateral (LAT) cervical spine. Four anatomical views had mean DI values outside of ±1, two of which required EIT update, one called for better patient positioning, and one required an image quality review to determine the needed technique adjustment.
Conclusion: We established an EI/DI performance baseline for one type of DR system at our institution. The study also provided an analysis framework for future ongoing exposure analysis.