Evaluation of dosimetric impact of lead ball marker artifacts in pelvic EBRT treatment planning

Abstract

Computed tomography (CT) imaging artifacts caused by the presence of high-density materials, such as lead ball fiducial markers, can introduce inaccuracies in Hounsfield unit (HU) values, which might lead to dose miscalculations in external beam radiation therapy (EBRT) planning. This study investigated the impact of these lead-marker-induced artifacts on dose calculations in pelvic EBRT plans. The pelvic region was chosen as the marker positioning was relatively uniform across patients, unlike in other regions. The primary aim was to quantify the differences in calculated dose distribution of artifact included and artifact-removed CT datasets, and to compare matrices between them to assess whether the observed differences are significant both statistically and clinically. Ten pelvic CT datasets were obtained from the existing datasets, each containing an original image series with visible lead markers and existing treatment plans. Each CT series was manually edited, and appropriate HU values were assigned to create an artifact-free version. Identical EBRT plans were applied to both CT sets using the VARIAN ARIA treatment planning system with the anisotropic analytical algorithm. Then, a plan comparison dose-volume histogram was generated for the artifact region and compared visually. Differences in key dose matrices (maximum, mean, median percentages) for the artifact structure averaged approximately 0.13%, 0.13% and 0.20%, respectively, across the dataset. According to the paired t-test analysis, there was no statistical significance across these parameters. However, in special cases where the clinical target volume (CTV) overlapped with the artifact region, more noticeable differences were observed. This indicates the potential for clinical relevance, highlighting the need for further study. The results highlight the importance of recognising and addressing high-density CT artifacts, especially the lead-ball artifact during pelvic EBRT planning, and the need for larger datasets as well as automated removal methods to enhance treatment accuracy and consistency.