Abstract
OBJECTIVE: Current age-based breast cancer screening protocols may not be optimally effective as they overlook mammographic density as a key risk factor. This study developed a personalized risk stratification model by analyzing age-specific mammographic density patterns to improve screening accuracy and reduce false-positive rates. METHODS: A cross-sectional analysis was performed on mammographic data from 2584 women aged 32-90 years from October 2023-December 2024. Breast Imaging Reporting and Data System (BI-RADS) density classifications were analyzed using polynomial regression and changepoint analysis to identify critical age thresholds. Four age-density clusters were derived, and a gradient boosting model was developed to evaluate predictive accuracy. RESULTS: The analysis identified three significant age thresholds (42.3, 51.7, and 65.2 years) where mammographic density patterns shifted. Four risk clusters were established, and the model achieved high predictive accuracy (Area Under the Curve [AUC] = 0.83). Simulations projected that personalized screening protocols could increase cancer detection by 14.7 % and reduce false positives by 9.7 % compared to traditional age-only approaches. CONCLUSIONS: Age-specific mammographic density screening offers a data-driven method to advance breast cancer prevention. It provides a framework for developing more effective screening policies that can decrease morbidity, supporting a shift toward risk-based screening as standard care.