Tube current limitation in CT automatic exposure control across two technology eras: a 15-year retrospective comparison spanning scanner replacement and team-based protocol optimization

Background Automatic exposure control (AEC) adapts CT tube current to patient attenuation, but only within the available tube current range. When the requested tube current falls outside that range, tube current plateaus at its maximum or minimum and image noise departs from the prescribed level. We compared the frequency and severity of tube current limitation in routine chest and abdomen-pelvis CT between 2007 and 2022, an interval spanning complete scanner replacement and the implementation of team-based CT protocol optimization. Methods This retrospective single-center study included routine chest (2007, n = 269; 2022, n = 291) and abdomen-pelvis (2007, n = 270; 2022, n = 469) examinations from GE scanners. Slice-level tube current, water-equivalent diameter (WED) and image noise were extracted from one axial series per examination. A percentile-based algorithm detected tube current plateaus, classified examinations as maximized (C Max ), minimized (C Min ), both (C Both ) or neither (C Optimized ), and quantified severity as the percentage of slices at the limit. Eras were compared at the examination level, with regression adjustment for WED and body region. Results Severe maximization (≥ 50% of slices at the maximum) decreased from 29.0% to 0.3% of chest and from 31.9% to 3.4% of abdomen-pelvis examinations (adjusted odds ratio 0.02, 95% CI 0.01–0.04); maximization throughout the examination decreased from 18.6% to 0% and from 16.7% to 1.3%. Any maximization decreased for chest (75.8% to 60.8%) but not for abdomen-pelvis (76.7% to 81.4%). Median noise did not increase (chest 5.08 vs. 4.44 HU; abdomen-pelvis 4.73 vs. 4.23 HU), although abdomen-pelvis patients were larger in 2022 (median WED 291 vs. 282 mm). Among scan`ners with the same detector width, the reduction in severe maximization persisted, whereas noise did not differ for chest (4.29 vs. 4.23 HU) and was slightly higher in 2022 for abdomen-pelvis (4.25 vs. 3.89 HU). Conclusions Severe, prolonged tube current maximization, concentrated in larger patients in 2007, was rare in 2022, whereas brief maximization remained common. Because scanner replacement and protocol optimization occurred together, these differences reflect their combined effect. Tube current limitation is a simple, transferable metric for monitoring AEC performance.