Optimization of a CT simulation protocol for breast cancer radiotherapy using response surface methodology

This study optimized a CT simulation protocol for breast cancer radiotherapy using Response Surface Methodology (RSM) to balance radiation dose and image quality. A Box-Behnken design evaluated three acquisition parameters: tube voltage (kVp), tube current-time product (mAs), and Sinogram Affirmed Iterative Reconstruction (SAFIRE) strength, measuring their effects on CT dose index (CTDIvol), image noise, and contrast-to-noise ratio (CNR). Scanning was performed on a Siemens SOMATOM Confidence scanner using a CIRS 610 phantom. Quadratic models developed for each response demonstrated high significance (p < 0.0001) and strong predictive power (R 2 > 0.98). Multi-objective optimization identified an optimal protocol (SAFIRE = 5, kVp = 120, mAs = 90) that reduced CTDIvol by 25%, decreased image noise by 60%, and improved CNR by 103% compared to the standard clinical protocol using filtered back projection. The application of RSM proved to be an effective approach for CT protocol optimization on the investigated CT system, enabling substantial dose reduction while improving image quality, thereby providing a quantitative basis for radiotherapy planning and protocol optimization.