Optimized low-kV spectrum of dual-energy CT equipped with high-kV tin filtration for electron density measurements.

PURPOSE

This paper describes the low-kV spectral optimization of dual-energy CT (DECT) equipped with high-kV tin filtration for the quantitative acquisition of electron density information, which is essential for treatment planning in radiotherapy. In addition, an analytical DECT image simulation was preliminarily performed to demonstrate the effectiveness of the optimized DECT with respect to the beam-hardening reduction.

METHODS

To optimize the low-kV spectrum of DECT, the author calculated the beam-hardening error, CT dose index, and tube loadings for a 50-cm diameter cylindrical water phantom with various combinations of filter materials, a range of thicknesses, and low-kV tube potentials. In addition, a single tube potential of 140 kV filtered by 0.4 mm tin (Sn) was employed for high-kV scanning, as is similar to the commercial implementation of the second-generation dual-source CT scanner. The optimized spectral parameters were then applied to the analytical DECT image simulation using two-dimensional fan-beam geometry for a virtual solid water phantom with 16 bodylike tissue inserts.

RESULTS

The author predicts that an optimal low-kV filtration would be 0.144-mm tungsten (W) at 90 kV, as it yields a minimal beam-hardening error with lower tube loadings and dose. The high-kV tube loading and dose obtained using the W filtration were 99 mAs and 2.2 mGy, respectively. These values are nearly equal to those obtained in the case of 2.5 mm Al at 100 kV (100 mAs and 2.3 mGy), which was regarded in this study as a reference filtration; however, the W filtration significantly reduced the beam-hardening error, from 9.5 to 1.4%. The corresponding low-kV tube loading (112 mAs) was five times greater than that of the reference case (21 mAs), but it was maintained at a certain practical level since the low-kV tube loading was comparable to the high-kV tube loading of the reference (100 mAs). The superiority of the beam-hardening reduction is reflected in the simulated images; for example, by the use of the W filter, the beam-hardening-induced deviation between the simulated and theoretical electron density values of cortical bone was reduced from 7.4 to 1.2% as compared with the reference filtration, even though no correction for beam hardening was performed.

CONCLUSIONS

In terms of beam hardening reduction, the DECT with the low-kV W filtration is more effective for the quantitative measurement of electron density within a practical limit of tube loadings and without additional dose.