Kenta Kitagawa1, Hitoshi Ikushima2, Motoharu Sasaki2, Shunsuke Furutani3, Takashi Kawanaka3, Akiko Kubo3, Chisato Tonoiso3, Takaharu Kudoh4, Yosuke Kano1, Akira Tsuzuki5. 1. Graduate School of Health Sciences, Tokushima University, Tokushima, 770-8503, Japan. 2. Department of Therapeutic Radiology, Institute of Biomedical Sciences, Tokushima University Graduate School, Tokushima, 770-8503, Japan. 3. Department of Radiology and Radiation Oncology, Institute of Biomedical Sciences, Tokushima University Graduate School, Tokushima, 770-8503, Japan. 4. Department of Oral and Maxillofacial Surgery, Institute of Biomedical Sciences, Tokushima University Graduate School, Tokushima, 770-8503, Japan. 5. Department of Radiological Technology, Kochi Medical School Hospital, Nankoku-shi, Kochi, 783-8505, Japan.
Abstract
PURPOSE: During treatment planning for head-and-neck volumetric-modulated arc therapy (VMAT), manual contouring of the metal artifact area of artificial teeth is done, and the area is replaced with water computed tomography (CT) values for dose calculation. This contouring of the metal artifact areas, which is performed manually, is subject to human variability. The purpose of this study is to evaluate and analyze the effect of inter-observer variation on dose distribution. METHODS: The subjects were 25 cases of cancer of the oropharynx for which VMAT was performed. Six radiation oncologists (ROs) performed metal artifact contouring for all of the cases. Gross tumor volume, clinical target volume, planning target volume (PTV), and oral cavity were evaluated. The contouring of the six ROs was divided into two groups, small and large groups. A reference RO was determined for each group and the dose distribution was compared with those of the other radiation oncologists by gamma analysis (GA). As an additional experiment, we changed the contouring of each dental metal artifact area, creating enlarged contours (L), reduced contours (S), and undrawn contours (N) based on the contouring by the six ROs and compared these structure sets. RESULTS: The evaluation of inter-observer variation showed no significant difference between the large and small groups, and the GA pass rate was 100%. Similar results were obtained comparing structure sets L and S, but in the comparison of structure sets L and N, there were cases with pass rates below 70%. CONCLUSIONS: The results show that the artificial variability of manual artificial tooth metal artifact contouring has little effect on the dose distribution of VMAT. However, it should be noted that the dose distribution may change depending on the contouring method in cases where the overlap between PTV and metal artifact areas is large.
PURPOSE: During treatment planning for head-and-neck volumetric-modulated arc therapy (VMAT), manual contouring of the metal artifact area of artificial teeth is done, and the area is replaced with water computed tomography (CT) values for dose calculation. This contouring of the metal artifact areas, which is performed manually, is subject to human variability. The purpose of this study is to evaluate and analyze the effect of inter-observer variation on dose distribution. METHODS: The subjects were 25 cases of cancer of the oropharynx for which VMAT was performed. Six radiation oncologists (ROs) performed metal artifact contouring for all of the cases. Gross tumor volume, clinical target volume, planning target volume (PTV), and oral cavity were evaluated. The contouring of the six ROs was divided into two groups, small and large groups. A reference RO was determined for each group and the dose distribution was compared with those of the other radiation oncologists by gamma analysis (GA). As an additional experiment, we changed the contouring of each dental metal artifact area, creating enlarged contours (L), reduced contours (S), and undrawn contours (N) based on the contouring by the six ROs and compared these structure sets. RESULTS: The evaluation of inter-observer variation showed no significant difference between the large and small groups, and the GA pass rate was 100%. Similar results were obtained comparing structure sets L and S, but in the comparison of structure sets L and N, there were cases with pass rates below 70%. CONCLUSIONS: The results show that the artificial variability of manual artificial tooth metal artifact contouring has little effect on the dose distribution of VMAT. However, it should be noted that the dose distribution may change depending on the contouring method in cases where the overlap between PTV and metal artifact areas is large.
Authors: Felix E Diehn; Gregory J Michalak; David R DeLone; Amy L Kotsenas; E Paul Lindell; Norbert G Campeau; Ahmed F Halaweish; Cynthia H McCollough; Joel G Fletcher Journal: Acta Radiol Open Date: 2017-11-26