A practical slice averaged image method for precise CT size specific dose estimates

Computed tomography (CT) is a widely used diagnostic tool, but variations in patient anatomy make accurate radiation dose estimation challenging. While size-specific dose estimates (SSDE) can serve as a more effective tool for dose prediction than conventional indices like CTDI_vol and DLP, calculating the water-equivalent diameter (D_w) for each slice is time-consuming and often impractical. To address this, we propose a novel method, the SSDE slice-averaged image (SSDE_SAI), which calculates D_w from a single image generated by averaging CT values across all slices. This approach captures anatomical variability while reducing calculation effort. We retrospectively analyzed CT data from 282 adult patients in three scan regions: chest, abdomen-pelvis, and chest-abdomen-pelvis (CAP). SSDE_SAI was compared with SSDE_center and the reference mean SSDE using regression analysis and root mean square error (RMSE). SSDE_SAI showed stronger agreement with mean SSDE than SSDE_center across all regions, achieving R² values up to 0.991 and lower RMSE. These results suggest that SSDE_SAI is a more advanced approach for dose prediction and may serve as a practical alternative for routine clinical use.