Effects of the difference in tube voltage of the CT scanner on dose calculation

Computed Tomography (CT) measures the attenuation coefficient of an object and converts the value assigned to each voxel into a CT number. In radiation therapy, CT number, which is directly proportional to the linear attenuation coefficient, is required to be converted to electron density for radiation dose calculation for cancer treatment. However, if various tube voltages were applied to take the patient CT image without applying the specific CT number to electron density conversion curve, the accuracy of dose calculation would be unassured. In this study, changes in CT numbers for different materials due to change in tube voltage were demonstrated and the dose calculation errors in percentage depth dose (PDD) and a clinical case were analyzed. The maximum dose difference in PDD from TPS dose calculation and Monte Carlo simulation were 1.3 % and 1.1 % respectively when applying the same CT number to electron density conversion curve to the 80 kVp and 140 kVp images. In the clinical case, the different CT number to electron density conversion curves from 80 kVp and 140 kVp were applied to the same image and the maximum differences in mean, maximum, and minimum doses were 1.1 %, 1.2 %, 1.0 % respectively at the central region of the phantom and 0.6 %, 0.9 %, 0.8 % respectively at the peripheral region of the phantom.

💡 Research Summary

The paper investigates how variations in CT scanner tube voltage (kVp) affect the conversion of CT numbers to electron density and consequently the accuracy of radiation dose calculations used in cancer radiotherapy. CT numbers (Hounsfield Units) are directly proportional to the linear attenuation coefficient of the material being imaged, and they must be transformed into electron density values for dose computation. Because the X‑ray spectrum changes with tube voltage, the same material can exhibit different CT numbers when scanned at different kVp settings. If a single CT‑to‑electron‑density conversion curve—typically derived for a specific voltage—is applied to images acquired at other voltages, the resulting electron density map may be inaccurate, leading to dose calculation errors.

The authors performed a systematic study using phantom materials that simulate various tissue types. Images were acquired at low voltage (80 kVp) and high voltage (140 kVp). For each material, the CT numbers were measured and compared across the two voltages. The data showed that CT numbers can shift by more than 5 % solely due to the voltage change, with low‑Z materials showing smaller shifts and high‑Z materials showing larger shifts.

To quantify the impact on dose calculation, the same CT‑to‑electron‑density conversion curve (derived for a reference voltage) was applied to both the 80 kVp and 140 kVp datasets. Percentage depth dose (PDD) curves were generated using a commercial treatment planning system (TPS) and compared with Monte Carlo (MC) simulations, which serve as a gold‑standard reference. The maximum discrepancy in PDD between the two voltage datasets was 1.3 % when the same conversion curve was used, while the difference between TPS and MC calculations was 1.1 %. These values indicate that the voltage‑induced CT number variation translates into a modest but non‑negligible dose error.

A clinical scenario was also examined. The same patient CT was reconstructed at both voltages, and two separate conversion curves—one appropriate for 80 kVp and one for 140 kVp—were applied to each dataset. An intensity‑modulated radiotherapy (IMRT) plan was then calculated for each case. In the central (high‑density) region of the phantom, the mean, maximum, and minimum dose differences between the two plans were 1.1 %, 1.2 %, and 1.0 % respectively. In the peripheral (low‑density) region, the corresponding differences were 0.6 %, 0.9 %, and 0.8 %. Although these deviations are relatively small, they become clinically relevant in high‑precision treatments such as stereotactic radiosurgery (SRS) or stereotactic body radiotherapy (SBRT), where sub‑percent accuracy is often required.

The study concludes that tube voltage influences CT numbers sufficiently to affect electron‑density mapping and dose calculation. While the observed errors (≤1.3 %) are within typical clinical tolerances, they highlight the importance of using voltage‑specific conversion curves or implementing robust correction algorithms when multi‑kVp imaging is employed. The authors recommend establishing separate CT‑to‑electron‑density calibration curves for each routinely used voltage, incorporating voltage verification into quality‑assurance (QA) protocols, and periodically validating TPS dose calculations against Monte Carlo simulations. By doing so, the radiotherapy community can maintain the high level of dosimetric accuracy required for modern, highly conformal treatment techniques.