Automation: The Future of Radiotherapy

Abstract

; the 80-degree collimator angle arc is fixed for all plans. The use of jaw tracking technique is chosen during the optimization process, which is commercially available. The two prescription dose levels of the planning target volumes were 70 Gy and 63 Gy in 35 fractions, using SIB. Results: The mean PTV 70 Gy and PTV 63 Gy doses were 65.83 Gy and 61.3 Gy for all plans respectively, extracted from the dose-volume histogram (DVH). Although the (60, 300 , and 80) plan shows slightly the lowest CI (0.96 and 0.94 for both PTV 70 Gy and PTV 63 Gy, respectively) and HI (0.06 and 0.14 for both PTV 70 Gy and PTV 63 Gy, respectively); the average values of conformity and homogeneity indexes were not significantly different for all plans. In addition, in terms of sparing of normal structures, the average DVHs showed that the maximum doses received by organs at risk were in the plan with highest collimator angles (60, 300 , and 80), and the minimum doses were in favor of the plan with the lowest collimator angles (15 , 345 , and 80). The averaged body integral doses were almost equivalent in all plans ranged between 130.38 Gy.L to 132.01 Gy.L. Conclusion: The dosimetric results for this study; indicated that the DVH parameters for the normal organs are dependent on the chosen collimator angles for all VMAT fields. As the collimator angle increases, the normal tissue doses slightly increase consequently. The PTVs HI and CI of all plans were almost the same, regardless of the collimator angles. The 80degree collimator angle arc gives additional sparing of OAR and more homogeneous PTV doses. Therefore, in order to improve the plan quality in NPC VMAT with SIB the planning physicist must select the optimal collimator angles that provide the best PTV coverage with that clinically acceptable OAR doses.

Development and validation of a checklist for use with automatically generated radiotherapy plans

“…[12][13][14][15][16][17][18][19][20][21] Automation can streamline the process, limiting the need for human interaction and decreasing the planning time. [22][23][24] While this lack of human input could limit human error, it could also decrease the error detection rate because of the lack of human review. Because of the different workflows used in automated contouring and treatment planning tools, the effectiveness of manual checklists in the physics review process, specifically in automated plans, should be evaluated.…”

“…New treatment planning tools automate aspects of the planning process, including contouring, planning, and quality assurance 12–21 . Automation can streamline the process, limiting the need for human interaction and decreasing the planning time 22–24 . While this lack of human input could limit human error, it could also decrease the error detection rate because of the lack of human review.…”