Click here to


Are you sure ?

Yes, do it No, cancel

Hospital-Wide Exposure Index Analysis for the Optimization of Exam-Specific Target Exposure Index

T Roberts*, L DeWeese , T Griglock , Oregon Health & Science Univ, Portland, OR


(Wednesday, 8/1/2018) 10:00 AM - 10:30 AM

Room: Exhibit Hall | Forum 8

Purpose: In 2008, IEC 62494 (AAPM TG-116) was published to standardize an Exposure Index with the goal of slowing dose creep resulting from decreased noise and adequate contrast with higher exposure in digital radiography (DR). It defined Exposure Index (EI) as an indicator for the incident air kerma on the DR detector and Deviation Index (DI) as a quality control tool for technologists representing kerma deviation from a Target Exposure Index (TEI). Non-exam-specific TEIs (body and extremity) are provided by vendors, but individual users must use exposure statistics to set TEIs appropriate for their practice. TG-116 (>±-3) and the upcoming TG-232 (>±2SD) provide guidelines for DI levels that potentially require action.

Methods: Data from more than 43,000 exposures from 11 DR rooms within a single hospital system was exported using Agfa’s Extended Dose Reporting. The exam types with greater than 100 exposures were analyzed in MatLab. TEIs were optimized by comparing the distribution of DIs calculated with both the median and geometric mean. The percentage of exposures within the target range and potentially requiring corrective action were compared to the percentage using the original TEI.

Results: The AP Chest view was optimized to a TEI of 410 from 250. The percentage of exams within the target range (≤±0.5) increased from 9.0% to 16.0%, and exams potentially requiring action (>±2SD) decreased from 10.1% to 5.5%. The lateral lumbar view had similar results (250 to 494, 5.6% to 19.6%, and 23.5% to 5.4%). The PA Landscape Chest view had a multi-modal distribution and had an increase in exams requiring attention (4.4% to 5.5%).

Conclusion: Exam-specific TEIs provide a better QC tool to technologists than a one-size-fits-all TEI. The median and geometric mean are adequate predictors of exam-specific TEI, but care should be taken when there is a multi-modal distribution.


Exposure Reduction, Digital Imaging, Quality Control


IM- X-ray: Quality Control

Contact Email