Exposure Indicator Drift in a New DR Room

The deviation index is defined as 10 × log10(receptor exposure ÷ target exposure), so a DI of +5 means 10^0.5, or about 3.2 times the target receptor exposure. That is a meaningful overexposure and explains dose creep, but DI describes exposure, not diagnostic quality. The images should not be repeated solely because of the DI: a diagnostic image with no motion, correct collimation, and correct positioning should be accepted and the technique corrected before the next patient, because repeating adds dose without improving diagnostic information. Repeat only if the image is nondiagnostic - motion, positioning error, inadequate collimation, or failure to include required anatomy. Also check whether the technologist used AEC or manual technique and whether the correct chamber or protocol was selected, since the DI pattern on three consecutive patients suggests a systematic problem rather than random error.

Likely causes fall into three groups. Equipment: AEC chamber calibration or incorrect chamber selection, generator kVp or mAs linearity drift, detector flat-field or gain calibration, grid damage or misalignment, or a misread SID indicator. Technique and process: a chart copied from a different room, wrong protocol or focal spot selected, body-habitus assumptions that no longer match the patient mix, or collimation that leaves the chamber partly outside the anatomy. Records: reject analysis by room, technologist, and exam type. Tests: flat-field uniformity and detector calibration check, AEC reproducibility and chamber-to-chamber consistency, kVp accuracy and mAs linearity, half-value layer, collimation and beam alignment, and SID verification. Repeat a phantom lateral lumbar exposure at the current chart settings and at corrected settings, and compare DI and noise. Document every measurement so the engineer and physicist can act on data rather than a guess.

First verify that the DI scale and the AEC are working, because correcting the chart while the AEC is miscalibrated will simply move the error. Then reduce receptor exposure by about a factor of three - for example, reducing mAs from 40 to roughly 13 at the same kVp - and reimage the phantom to confirm DI near 0 with acceptable noise. Because reducing mAs increases quantum mottle, evaluate the phantom and then a small series of patients rather than assuming the lower number is automatically better; if noise becomes limiting, reduce kVp or accept a DI closer to +1 to +2. Keep kVp changes deliberate, since kVp governs contrast and the radiologist already called the images flat. Update the chart for all projections in that room, not just the lateral lumbar, and recheck after 20–30 patients. Monitor DI distribution and the reject rate by room weekly, and re-audit after any service visit.