Purpose: To evaluate the performance of the metal artifact reduction algorithm in radiation therapy.
Methods: The iterative metal artifact reduction (iMAR) algorithm combines beam hardening correction (iBHC) and sinogram inpainting. To quantify the performance of the iMAR, an electron density phantom (Gammex RMI 465) was scanned with and without titanium and aluminum inserts with and without the application of iMAR. A clinically-used pelvis protocol scan was used (Siemens SOMATOM go.Sim). A quantitative evaluation was done using the signal-to-noise ratio (SNR) of uncorrected, iMAR, and iMAR with iBHC images. Locations were selected near the area of streaking artifacts. The DVH for a 6 MV static beam was used for dosimetric comparison (AAA algorithm Varian, Eclipse 15.6). A clinical case with dental fillings was reviewed to verify that the recommended algorithm ‘iMAR Dental fillings’ can be considered as an optimal algorithm for the case. Two algorithms, ‘Extremity implants’ and ‘Dental fillings’ were used to perform image reconstruction to observe the differences.
Results: The SNRs obtained with iMAR appear higher compared to those with the uncorrected for high-density materials: 25.5% and 14.1% for titanium and aluminum, respectively. With iMAR + iBHC, the SNR increases by 43.1% and 32.9%. In the selected ROI, the dosimetric difference was 0.2% of the dose in the iMAR case and there was a 0.3% increase in dosimetry with iMAR + iBHC. For a clinical case, a significant difference in image quality between ‘Extremity implants’ and ‘Dental fillings’ algorithms, was observed. Using the ‘Dental fillings’ algorithm, an additional artifact was found outside of the high-density region.
Conclusion: This work indicates the iMAR with iBHC significantly improves images obscured by metal artifacts and the accuracy of calculated dose distributions. Further evaluation is needed to substantiate the results to apply different reconstruction parameters for various high-density materials.
CT, Image Artifacts, Radiation Therapy