Room: Exhibit Hall
Purpose: The presence of large metallic implants causes severe artefacts in CT images. These metal artefacts affect negatively the treatment planning by causing difficulties to delineate the target volume and by reducing the dose calculation accuracy. The Philips metal artefact reduction algorithm for orthopaedic implants, O-MAR, improves image quality and visualization of anatomy by reducing the effect of metal on the images. This study evaluated the treatment planning dose accuracy on O-MAR CT data.
Methods: A commercial metal hip prosthesis made of titanium and stainless steel was implanted in a custom made wax cylinder plug. The ArcCheck phantom was scanned with this wax cylinder plug on a Philips Brilliance Big Bore CT scanner. Dosimetric studies were carried out for 6MV, 10MV and 18MV on uncorrected (Uncorr), manually artefact corrected (Corr) with assigned density, and O-MAR corrected (OMAR) CT data set in Varian Eclipse TPS using static and arc open fields.
Results: ArcCheck gamma analysis between measured and planned dose fluence showed better pass rate for dose calculated on Corr data set for static fields. The effect is more for 6MV as opposed to 10MV and 18MV. No such difference was observed for arc fields. Planned doses on Uncorr, Corr and OMAR data sets were compared against each other and gamma analysis were performed. The gamma pass rate with 1%, 1mm criterion for Corr vs Uncorr were in the range of 88.9%-93.3% (static fields) and 95.3%-99.5% (arc fields), and for Corr vs OMAR were in the range of 90.5%-93.9% (static fields) and 94.8%-98.4% (arc fields). But the gamma pass rate for OMAR vs Uncorr were in the range 97.6%-97.9% (static fields) and 98.5%-98.8% (arc fields).
Conclusion: This study indicated that dose calculation on manually artefact corrected images with the help of O-MAR data can give us the highest dose calculation accuracy.