Characterizing cumulative lens dose in recurrent CT: A Large-Scale Single-Center cohort study of drivers.

OBJECTIVES

To quantify the distribution and evolution patterns of the annual cumulative lens dose (ACLD) in recurrent CT examinations and identify the key drivers of extreme radiation burdens.

MATERIALS AND

METHODS

This retrospective cohort study included 118,126 patients who underwent recurrent CT examinations (≥ 2 scans within a 12-month rolling window) at a single center between January 2021 and December 2025. A log-transformed ordinary least squares (OLS) regression model with interaction terms was employed to evaluate ACLD trends. Furthermore, specific attribution analyses were conducted across high-ACLD subgroups-stratified by clinical thresholds of ≥ 100, ≥250, and ≥ 500 mGy-to evaluate the impact of scan frequency, anatomical site, scan type, and clinical indications on extreme dose accumulation.

RESULTS

In the total recurrent cohort, the observed median ACLD was 60.637 mGy (IQR: 0.55-105.51 mGy). In the multivariable interaction model, the predicted conditional geometric mean ACLD for the reference profile (a 52-year-old male undergoing non-contrast head CT for neurological indications) was 68.975 mGy. Across the overall cohort, ACLD accumulation demonstrated "scatter-driven proportional growth," with distal sites (chest, abdomen, extremities) showing significantly higher effective geometric mean ratios (GMRs > 1.6) per additional scan. Conversely, in the high-dose subgroups (≥ 250 mGy), the marginal contribution of trunk scatter saturated (GMRs ≈1.0). At extreme dose levels (≥ 500 mGy), ACLD accumulation was overwhelmingly driven by direct head exposures for neurological emergencies, completely neutralizing the influence of baseline physiological characteristics like sex.

CONCLUSION

Cumulative lens dose growth evolves dynamically from scatter-driven proportional increases at low baselines to direct exposure-dominated absolute accumulation at high burdens. Implementing organ-specific dose monitoring is crucial for high-frequency imaging patients.