[en] Brachytherapy devices and software are designed to last for a certain period of time. Due to a number of considerations, such as material factors, wear-and-tear, backwards compatibility, and others, they all reach a date when they are no longer supported by the manufacturer. Most of these products have a limited duration for their use, and the information is provided to the user at time of purchase. Because of issues or concerns determined by the manufacturer, certain products are retired sooner than the anticipated date, and the user is immediately notified. In these situations, the institution is facing some difficult choices: remove these products from the clinic or perform tests and continue their usage. Both of these choices come with a financial burden: replacing the product or assuming a potential medicolegal liability. This session will provide attendees with the knowledge and tools to make better decisions when facing these issues. Learning Objectives: Understand the meaning of “end-of-life or “life expectancy” for brachytherapy devices and software Review items (devices and software) affected by “end-of-life” restrictions Learn how to effectively formulate “end-of-life” policies at your institution Learn about possible implications of “end-of-life” policy Review other possible approaches to “end-of-life” issue

[en] Purpose: To evaluate geometric and dosimetric uncertainties of CT-CBCT deformable image registration (DIR) algorithms using digital phantoms generated from real patients. Methods: We selected ten H&N cancer patients with adaptive IMRT. For each patient, a planning CT (CT1), a replanning CT (CT2), and a pretreatment CBCT (CBCT1) were used as the basis for digital phantom creation. Manually adjusted meshes were created for selected ROIs (e.g. PTVs, brainstem, spinal cord, mandible, and parotids) on CT1 and CT2. The mesh vertices were input into a thin-plate spline algorithm to generate a reference displacement vector field (DVF). The reference DVF was applied to CBCT1 to create a simulated mid-treatment CBCT (CBCT2). The CT-CBCT digital phantom consisted of CT1 and CBCT2, which were linked by the reference DVF. Three DIR algorithms (Demons, B-Spline, and intensity-based) were applied to these ten digital phantoms. The images, ROIs, and volumetric doses were mapped from CT1 to CBCT2 using the DVFs computed by these three DIRs and compared to those mapped using the reference DVF. Results: The average Dice coefficients for selected ROIs were from 0.83 to 0.94 for Demons, from 0.82 to 0.95 for B-Spline, and from 0.67 to 0.89 for intensity-based DIR. The average Hausdorff distances for selected ROIs were from 2.4 to 6.2 mm for Demons, from 1.8 to 5.9 mm for B-Spline, and from 2.8 to 11.2 mm for intensity-based DIR. The average absolute dose errors for selected ROIs were from 0.7 to 2.1 Gy for Demons, from 0.7 to 2.9 Gy for B- Spline, and from 1.3 to 4.5 Gy for intensity-based DIR. Conclusion: Using clinically realistic CT-CBCT digital phantoms, Demons and B-Spline were shown to have similar geometric and dosimetric uncertainties while intensity-based DIR had the worst uncertainties. CT-CBCT DIR has the potential to provide accurate CBCT-based dose verification for H&N adaptive radiotherapy. Z Shen: None; K Bzdusek: an employee of Philips Healthcare; S Koyfman: None; P Xia: received research grants from Philips Healthcare and Siemens Healthcare

[en] Purpose: To introduce methods to analyze Deformable Image Registration (DIR) and identify regions of potential DIR errors. Methods: DIR Deformable Vector Fields (DVFs) quantifying patient anatomic changes were evaluated using the Jacobian determinant and the magnitude of DVF curl as functions of tissue density and tissue type. These quantities represent local relative deformation and rotation, respectively. Large values in dense tissues can potentially identify non-physical DVF errors. For multiple DVFs per patient, histograms and visualization of DVF differences were also considered. To demonstrate the capabilities of methods, we computed multiple DVFs for each of five Head and Neck (H'N) patients (P1–P5) via a Fast-symmetric Demons (FSD) algorithm and via a Diffeomorphic Demons (DFD) algorithm, and show the potential to identify DVF errors. Results: Quantitative comparisons of the FSD and DFD registrations revealed <0.3 cm DVF differences in >99% of all voxels for P1, >96% for P2, and >90% of voxels for P3. While the FSD and DFD registrations were very similar for these patients, the Jacobian determinant was >50% in 9–15% of soft tissue and in 3–17% of bony tissue in each of these cases. The volumes of large soft tissue deformation were consistent for all five patients using the FSD algorithm (mean 15%±4% volume), whereas DFD reduced regions of large deformation by 10% volume (785 cm³) for P4 and by 14% volume (1775 cm³) for P5. The DFD registrations resulted in fewer regions of large DVF-curl; 50% rotations in FSD registrations averaged 209±136 cm³ in soft tissue and 10±11 cm³ in bony tissue, but using DFD these values were reduced to 42±53 cm³ and 1.1±1.5 cm³, respectively. Conclusion: Analysis of Jacobian determinant and curl as functions of tissue density can identify regions of potential DVF errors by identifying non-physical deformations and rotations. Collaboration with Phillips Healthcare, as indicated in authorship

[en] Purpose: To generate virtual phantoms with clinically relevant deformation and use them to objectively evaluate geometric and dosimetric uncertainties of deformable image registration (DIR) algorithms. Methods: Ten lung cancer patients undergoing adaptive 3DCRT planning were selected. For each patient, a pair of planning CT (pCT) and replanning CT (rCT) were used as the basis for virtual phantom generation. Manually adjusted meshes were created for selected ROIs (e.g. PTV, lungs, spinal cord, esophagus, and heart) on pCT and rCT. The mesh vertices were input into a thin-plate spline algorithm to generate a reference displacement vector field (DVF). The reference DVF was used to deform pCT to generate a simulated replanning CT (srCT) that was closely matched to rCT. Three DIR algorithms (Demons, B-Spline, and intensity-based) were applied to these ten virtual phantoms. The images, ROIs, and doses were mapped from pCT to srCT using the DVFs computed by these three DIRs and compared to those mapped using the reference DVF. Results: The average Dice coefficients for selected ROIs were from 0.85 to 0.96 for Demons, from 0.86 to 0.97 for intensity-based, and from 0.76 to 0.95 for B-Spline. The average Hausdorff distances for selected ROIs were from 2.2 to 5.4 mm for Demons, from 2.3 to 6.8 mm for intensity-based, and from 2.4 to 11.4 mm for B-Spline. The average absolute dose errors for selected ROIs were from 0.2 to 0.6 Gy for Demons, from 0.1 to 0.5 Gy for intensity-based, and from 0.5 to 1.5 Gy for B-Spline. Conclusion: Virtual phantoms were modeled after patients with lung cancer and were clinically relevant for adaptive radiotherapy treatment replanning. Virtual phantoms with known DVFs serve as references and can provide a fair comparison when evaluating different DIRs. Demons and intensity-based DIRs were shown to have smaller geometric and dosimetric uncertainties than B-Spline. Z Shen: None; K Bzdusek: an employee of Philips Healthcare; J Greskovich: None; P Xia: received research grants from Philips Healthcare and Siemens Healthcare

[en] Purpose: We propose a novel data-driven method to predict the achievability of clinical objectives upfront before invoking the IMRT optimization. Methods: A new metric called “Geometric Complexity (GC)” is used to estimate the achievability of clinical objectives. Here, GC is the measure of the number of “unmodulated” beamlets or rays that intersect the Region-of-interest (ROI) and the target volume. We first compute the geometric complexity ratio (GCratio) between the GC of a ROI (say, parotid) in a reference plan and the GC of the same ROI in a given plan. The GCratio of a ROI indicates the relative geometric complexity of the ROI as compared to the same ROI in the reference plan. Hence GCratio can be used to predict if a defined clinical objective associated with the ROI can be met by the optimizer for a given case. Basically a higher GCratio indicates a lesser likelihood for the optimizer to achieve the clinical objective defined for a given ROI. Similarly, a lower GCratio indicates a higher likelihood for the optimizer to achieve the clinical objective defined for the given ROI. We have evaluated the proposed method on four Head and Neck cases using Pinnacle3 (version 9.10.0) Treatment Planning System (TPS). Results: Out of the total of 28 clinical objectives from four head and neck cases included in the study, 25 were in agreement with the prediction, which implies an agreement of about 85% between predicted and obtained results. The Pearson correlation test shows a positive correlation between predicted and obtained results (Correlation = 0.82, r2 = 0.64, p < 0.005). Conclusion: The study demonstrates the feasibility of the proposed method in head and neck cases for predicting the achievability of clinical objectives with reasonable accuracy.

[en] Purpose: Database dose predictions and a commercial autoplanning engine both improve treatment plan quality in different but complimentary ways. The combination of these planning techniques is hypothesized to further improve plan quality. Methods: Four treatment plans were generated for each of 10 head and neck (HN) and 10 prostate cancer patients, including Plan-A: traditional IMRT optimization using clinically relevant default objectives; Plan-B: traditional IMRT optimization using database dose predictions; Plan-C: autoplanning using default objectives; and Plan-D: autoplanning using database dose predictions. One optimization was used for each planning method. Dose distributions were normalized to 95% of the planning target volume (prostate: 8000 cGy; HN: 7000 cGy). Objectives used in plan optimization and analysis were the larynx (25%, 50%, 90%), left and right parotid glands (50%, 85%), spinal cord (0%, 50%), rectum and bladder (0%, 20%, 50%, 80%), and left and right femoral heads (0%, 70%). Results: All objectives except larynx 25% and 50% resulted in statistically significant differences between plans (Friedman’s χ"2 ≥ 11.2; p ≤ 0.011). Maximum dose to the rectum (Plans A-D: 8328, 8395, 8489, 8537 cGy) and bladder (Plans A-D: 8403, 8448, 8527, 8569 cGy) were significantly increased. All other significant differences reflected a decrease in dose. Plans B-D were significantly different from Plan-A for 3, 17, and 19 objectives, respectively. Plans C-D were also significantly different from Plan-B for 8 and 13 objectives, respectively. In one case (cord 50%), Plan-D provided significantly lower dose than plan C (p = 0.003). Conclusion: Combining database dose predictions with a commercial autoplanning engine resulted in significant plan quality differences for the greatest number of objectives. This translated to plan quality improvements in most cases, although special care may be needed for maximum dose constraints. Further evaluation is warranted in a larger cohort across HN, prostate, and other treatment sites. This work is supported by Philips Radiation Oncology Systems.

[en] This study evaluates the performance and planning efficacy of the Auto-Planning (AP) module in the clinical version of Pinnacle 9.10 (Philips Radiation Oncology Systems, Fitchburg, WI, USA). Twenty automated intensity-modulated radiotherapy (IMRT) plans were compared with the original manually planned clinical IMRT plans from patients with oropharyngeal cancer. Auto-Planning with IMRT offers similar coverage of the planning target volume as the original manually planned clinical plans, as well as better sparing of the contralateral parotid gland, contralateral submandibular gland, larynx, mandible, and brainstem. The mean dose of the contralateral parotid gland and contralateral submandibular gland could be reduced by 2.5 Gy and 1.7 Gy on average. The number of monitor units was reduced with an average of 143.9 (18%). Hands-on planning time was reduced from 1.5-3 h to less than 1 h. The Auto-Planning module was able to produce clinically acceptable head and neck IMRT plans with consistent quality. (orig.)