Analysis of Possibility of Using an Algorithm for Correcting Metal Artifacts in CT-Images for Radiation Therapy Planning

Purpose: To quantitatively evaluate reconstruction software algorithms in combination with algorithm O-MAR for correcting metal artifacts in CT-images and explore the potential of using O-MAR for radiation therapy planning tasks. Material and methods: A quantitative assessment of the O-MAR algorithm on CT scans of a 20 cm diameter cylindrical phantom with a hip joint implant in the center was performed. Test tubes with different concentrations of potassium hydrogen phosphate (K2HO4×3H2O) were placed around the implant. The evaluation parameters used standard deviation (SD) of the region of interest (ROI) density in HU units and the calculation of the degree of susceptibility to artifacts (P).The calculation of absorbed dose in the phantom was performed on the Eclipse 17.0 planning station, using the AAA (Analytical Anisotropic Algorithm) calculation algorithm. Results: Calculations of the degree of artifact susceptibility showed that the minimum mean noise value was observed for the iMR series in combination with O-MAR (31.6 ± 45.5 HU) and the maximum for FBP (16) without O-MAR (77.0 ± 31.1 HU). As a result of comparison of CT studies with/without O-MAR, the average calculated difference in absorbed dose for all control points is 0.33±1.68 % and 0.42±1.38 % in the presence of implant for FBP and iMR modes, respectively. However, for both modes the difference was 3.22 % for the artifact zone (dark spot). Conclusion: It is shown that the use of the O-MAR algorithm reduces the distorted values of X-ray density that arose as a result of the presence of an implant in CT studies. The calculation of the absorbed dose for the artifact zone (dark spot) shows a decrease in the uncertainty of the dose calculation in O-MAR-corrected studies.