A Quick, User-Friendly and Interactive Approach for High-Dose-Rate and Pulsed-Dose-Rate Brachytherapy Treatment Planning: Enhanced Geometric Optimization - Interactive Inverse Planning

PurposeManual graphical optimization of stepping source (HDR and PDR) brachytherapy treatment plans is time consuming and requires an experienced user. We aim to accelerate and simplify this process. New treatment planning tools were developed for this purpose and evaluated on 14 prostate cases.Materials and MethodsTwo main tools were developed: EGO (Enhanced Geometric Optimization) and IIP (Interactive Inverse Planning). EGO creates a dose distribution, with a maximum value for the homogeneity parameter QI (Quality Index). This distribution can be used as a starting point for the interactive part: IIP. Within IIP, lowest and highest doses on the surfaces of all organs are shown. The user can interactively set new values for these lowest and highest doses, and the program adjusts the dwell times in order to change the dose limits to the desired level. Our last fourteen prostate cases were selected to test the new software. These cases had previously been planned by an experienced treatment planner using graphical optimization. These plans are considered as the benchmark and took between 30 and 45 minutes to make. Ten minutes were given to an inexperienced treatment planner to create a new treatment plan with EGO+IIP that was comparable or better than the benchmark plan. The following dose parameters were evaluated: prostate V100 (%) and D90 (cGy), Rectum V80 (cc), Urethra V125 (cc), Bladder V100 (cc) and TRAK ([email protected]). The planning constraints corresponding to these parameters were: pr-V100>95%, re-V80<2cc, ur-V125<1cc and bl-V100<1cc.ResultsTable 1Mean DVH parameters and their interquartile range (IQR) for the manually optimized plans and the plans created with EGO and IIP. Values displayed in bold are significantly improved as compared to the benchmark (graphical optimization).Pr-V100 (%)Pr-D90 (cGy)TRAK (cGy cm)Re-V80 (cc)Ur-V125 (cc)Bl-V100 (cc)Mean Graph. opt.94.02129.54610.930.590.39IQR Graph. opt.93.3-97.6129.6-137.4349-5360.03-1.420.25-0.950.13-0.36Mean EGO + IIP93.81130.44650.730.480.32IQR EGO + IIP92.5-96.9126.3-138.2361-5310.03-1.050.07-0.880.04-0.36 Open table in a new tab ConclusionsWith our newly developed method EGO+IIP a user with limited experience can create treatment plans comparable to or better than the graphically optimized plans. The main advantages are 1) the short amount of time needed and 2) the ease of optimization. PurposeManual graphical optimization of stepping source (HDR and PDR) brachytherapy treatment plans is time consuming and requires an experienced user. We aim to accelerate and simplify this process. New treatment planning tools were developed for this purpose and evaluated on 14 prostate cases. Manual graphical optimization of stepping source (HDR and PDR) brachytherapy treatment plans is time consuming and requires an experienced user. We aim to accelerate and simplify this process. New treatment planning tools were developed for this purpose and evaluated on 14 prostate cases. Materials and MethodsTwo main tools were developed: EGO (Enhanced Geometric Optimization) and IIP (Interactive Inverse Planning). EGO creates a dose distribution, with a maximum value for the homogeneity parameter QI (Quality Index). This distribution can be used as a starting point for the interactive part: IIP. Within IIP, lowest and highest doses on the surfaces of all organs are shown. The user can interactively set new values for these lowest and highest doses, and the program adjusts the dwell times in order to change the dose limits to the desired level. Our last fourteen prostate cases were selected to test the new software. These cases had previously been planned by an experienced treatment planner using graphical optimization. These plans are considered as the benchmark and took between 30 and 45 minutes to make. Ten minutes were given to an inexperienced treatment planner to create a new treatment plan with EGO+IIP that was comparable or better than the benchmark plan. The following dose parameters were evaluated: prostate V100 (%) and D90 (cGy), Rectum V80 (cc), Urethra V125 (cc), Bladder V100 (cc) and TRAK ([email protected]). The planning constraints corresponding to these parameters were: pr-V100>95%, re-V80<2cc, ur-V125<1cc and bl-V100<1cc. Two main tools were developed: EGO (Enhanced Geometric Optimization) and IIP (Interactive Inverse Planning). EGO creates a dose distribution, with a maximum value for the homogeneity parameter QI (Quality Index). This distribution can be used as a starting point for the interactive part: IIP. Within IIP, lowest and highest doses on the surfaces of all organs are shown. The user can interactively set new values for these lowest and highest doses, and the program adjusts the dwell times in order to change the dose limits to the desired level. Our last fourteen prostate cases were selected to test the new software. These cases had previously been planned by an experienced treatment planner using graphical optimization. These plans are considered as the benchmark and took between 30 and 45 minutes to make. Ten minutes were given to an inexperienced treatment planner to create a new treatment plan with EGO+IIP that was comparable or better than the benchmark plan. The following dose parameters were evaluated: prostate V100 (%) and D90 (cGy), Rectum V80 (cc), Urethra V125 (cc), Bladder V100 (cc) and TRAK ([email protected]). The planning constraints corresponding to these parameters were: pr-V100>95%, re-V80<2cc, ur-V125<1cc and bl-V100<1cc. ResultsTable 1Mean DVH parameters and their interquartile range (IQR) for the manually optimized plans and the plans created with EGO and IIP. Values displayed in bold are significantly improved as compared to the benchmark (graphical optimization).Pr-V100 (%)Pr-D90 (cGy)TRAK (cGy cm)Re-V80 (cc)Ur-V125 (cc)Bl-V100 (cc)Mean Graph. opt.94.02129.54610.930.590.39IQR Graph. opt.93.3-97.6129.6-137.4349-5360.03-1.420.25-0.950.13-0.36Mean EGO + IIP93.81130.44650.730.480.32IQR EGO + IIP92.5-96.9126.3-138.2361-5310.03-1.050.07-0.880.04-0.36 Open table in a new tab ConclusionsWith our newly developed method EGO+IIP a user with limited experience can create treatment plans comparable to or better than the graphically optimized plans. The main advantages are 1) the short amount of time needed and 2) the ease of optimization. With our newly developed method EGO+IIP a user with limited experience can create treatment plans comparable to or better than the graphically optimized plans. The main advantages are 1) the short amount of time needed and 2) the ease of optimization.