Click here to


Are you sure ?

Yes, do it No, cancel

Implementation of Flag Criteria in Analysis of Digital Radiography Dose Statistics

K Scilla1*, K Hulme2 , R Fisher3 , X Li2 , (1) Cleveland State University, Cleveland, OH, (2) The Cleveland Clinic, Beachwood, OH, (3) MetroHealth, Cleveland, OH


(Monday, 7/15/2019) 3:45 PM - 4:15 PM

Room: Exhibit Hall | Forum 8

Purpose: A retrospective analysis of a year’s worth of exposure index (EI) data was conducted to evaluate the state of practice among several radiography workstations in a large enterprise after a DR panel retrofit, with the goal of establishing clinically-appropriate target EI (TEI) values and identifying inconsistencies in practice.

Methods: One year after installation of wireless DR panels (Agfa-DR14s), exposure data was exported, yielding 150,269 dose statistics for analysis. Values for the mean and standard deviation of DIref were calculated as a function of exam group, anatomical view, and workstation. Flag criteria were established in order to determine which exam groups, anatomical views, and workstations needed further review.

Results: Aggregate data for C-Spine exam group was flagged for review due to SD(DIref)>2.5, prompting further analysis at the view level. Histograms were generated for AP and LAT c-spine, the latter having a very broad distribution. LAT c-spine and LAT EXT C-spine were both flagged for SD(DIref)>2.5 and negative kurtosis.The next step was a workstation-specific analysis of the LAT views. Workstations A1, A2, and D2 all had similar distributions and statistics, while E1 was an outlier with a narrower distribution centered at a different point than the others. In order to identify the root cause of these inconsistencies, the lead technologists at each site were contacted regarding their techniques. Different parameters were in use at each site, with site E1 using considerably higher mAs.

Conclusion: From this analysis, it was shown that for the C-Spine exam group, a single TEI value is inappropriate, as there is too much variability between views in this group. Similarly, it was determined that the lateral C-spine views displayed inconsistencies and a TEI could not be established for these views at this time. Implementing these flag criteria identified these inconsistencies so that they can be addressed.


Not Applicable / None Entered.


Not Applicable / None Entered.

Contact Email