N Sheth

SU‐GG‐I‐103: Comparison of Model‐Based Segmentation Systems for Contouring of Male Pelvic Structures

Purpose: Manual delineation of structures for radiation therapytreatment planning often is a laborious and time consuming process, particularly with IMRT, adaptive planning, and 4D CT data sets. In this study, two model‐based segmentation (MBS) systems are compared and evaluated for contouring of structures in the male pelvis. Method and Materials: Two commercially available MBS systems, Oncentra (Nucletron, Veenendaal NL) and Pinnacle (Philips Medical Systems, Madison WI) are used to delineate structures of the male pelvis for radiation therapytreatment planning. Five patients undergoing prostate IMRTtreatment were randomly selected and contoured in Pinnacle using non‐MBS tools; this is considered the benchmark. Contours were also generated using both MBS systems and subsequently corrected using manual non‐MBS tools that are similar to both Oncentra and Pinnacle. Contouring times and volumes for both MBS systems were compared and evaluated against the benchmark non‐MBS contours. Results: The average time to contour male pelvic structures with non‐MBS tools was 14.93 minutes. Using MBS, this was improved by 25.0% in Oncentra and 30.1% in Pinnacle. The mean reduction in time contouring using MBS with Oncentra and Pinnacle, respectively, was 2.8%±24.0 and 10.5%±15.5 for the prostate, 18.5%±24.0 and 25.5%±17.1 for the bladder, 2.6%±12.2 and 7.2% ±8.6 for the rectum, and 42.1%±15.1 and 45.4%±13.4 for both femurs. Using the MBS technique in Oncentra and Pinnacle, 28.9% ±8.0 and 31.6%±5.5 of the respective total contouring time was spent generating the MBS contours and the remainder of editing with non‐MBS tools. Contour volumes were similar for the benchmark and the MBS systems. Conclusion: Both MBS systems require manual editing of auto generated contours to better match the benchmarks. Nevertheless, the MBS systems resulted in similar time savings over utilizing only non‐MBS tools when delineating structures of the male pelvis for radiation therapytreatment planning.

SU‐E‐T‐378: Dosimetry Comparison of VMAT and Tomotherapy Plans with Dose Painting in Brain Metastases

Purpose: To compare dosimetric results of VMAT and Tomotherapy dose painting techniques with hippocampus sparing in brain metastases. Methods: A patient with multiple brain metastases was entered into this dosimetry study. Pinnacle VMAT SmartArc™ plan was executed in version 9.2 while setting the same dose constraints in Tomotherapy planning version 4.04. The clinical goal is to perform simultaneous in‐field boost (SIB) with whole brain for 32.5 Gy while delivery extra dose per fraction to the brain metastatic sites to 63 Gy. The mean dose to each of the hippocampus was prescribed to less than 6 Gy. Dosimetric analysis was performed on the 3DVH modules by Sun Nuclear Corp™. The ArcCheck™ platform was also utilized to measure the 3%/3mm gamma passing rate. Results: Clinical planning criteria for GTV coverage were all satisfied, while Tomotherapy presents better hippocampus sparing compared both to the VMAT technique (mean: 5.29/10.05 Gy on left and 5.18/9.90 Gy on right, max:7.19/17.82 Gy on left and 7.58/20.79 Gy on right). Tomotherapy has inferior chiasm sparing with mean 32.26 Gy compared to VMAT planning 29.16 Gy. However, the dose uniformity of Tomotherapy has proven to be superior in all lesions with average mean dose 63.63 Gy with respect to 64.33 Gy; max dose of 102.2% on Tomotherapy and 105.3% on VMAT planning. ArcCheck passing rate for Tomotherapy is 98.9% with 3%/3mm gamma settings. Conclusions: Tomotherapy planning is considered a gold standard for arc‐based treatment. From the planning comparison, with VMAT delivery, achieving 6Gy dose sparing to the hippocampus presents great challenge compared to the Tomotherapy planning and delivery. One major reason is that the VMAT MLC speed cannot generate enough fluence patterns to produce the extremely high dose gradient. ArcCheck measurements for VMAT and Tomotherapy represented errors induced from the combination of TPS dose calculation algorithm and beam‐delivery inaccuracy.

SU‐E‐J‐34: Influence of Prone versus Supine Patient Position on Localization with Image Guided Radiotherapy of Prostate Cancer

PURPOSE To retrospectively assess the influence of prone versus supine patient position on setup shifts and target margins for image guided radiotherapy (IGRT) of prostate cancer. METHODS Ninety prostate cancer patients were treated at our institution between September 2006 and June2011 with IGRT utilizing daily megavoltage cone beam; 52 patients were prone receiving 1577 fractions and 38 supine receiving 1035 fractions. Patients were setup to skin tattoos, imaged with megavoltage cone beam, then moved to treatment position by the alignment of three intraprostatic fiducials with lateral, longitudinal, and vertical shifts. The magnitude of the daily shift is root sum of squares of the translations. RESULTS Patients positioned prone had a mean daily shift magnitude 1.4 mm greater than supine: 7.5 ± 4.0 mm vs. 6.1 ± 2.9 mm (p ≪ 0.001). The mean magnitude of the daily shift for prone vs. supine respectively were 3.7 ± 3.1 mm vs. 2.6 ± 2.3 mm laterally, 3.5 ± 3.0 mm vs. 3.3 ± 2.6 mm longitudinally, and 3.9 ± 3.5 mm vs. 3.1 ± 2.4 mm vertically. The percentage of daily shifts within our intuitions planning target volume (PTV) margin (7 mm posterior and 10 mm for all other directions) for prone vs. supine respectively were 97.3% vs. 99.5% laterally, 97.1% vs. 98.2% longitudinally, and 93.1% vs. 96.8% vertically. CONCLUSION When prostate cancer patients were setup for IGRT utilizing daily megavoltage cone beam, the daily shifts from the skin tattoos to intraprostatic fiducials were greater on average if positioned prone vs.supine. Without IGRT, part of the prostate would be outside of the PTV for more treatments when positioned prone rather than supine.

SU‐E‐T‐428: Class Solution for Prostate High Dose Rate Brachytherapy with Inverse Planning Simulated Annealing

PURPOSE To develop and validate a class solution for inverse planning simulated annealing (IPSA) with CT based prostate high dose rate brachytherapy (HDR). METHODS Between November 2008 and November 2011, our institution treated 40 prostate cancer patients with HDR in 7 Gy fractions followed by external beam radiotherapy. The HDR treatments were planned with Nucletron Oncentra using manual graphical optimization (GO). Plans were optimized to the following clinical goals: = 95% of prostate volume received 7 Gy, < 1 cc of rectum received 5.6 Gy, < 0.1 cc of rectum received 6.3 Gy, and < 0.01 cc of urethra received 8.75 Gy. New plans were manually customized using IPSA (MC-IPSA) for each patient to match prostate coverage by the prescription dose to within ± 1% of the GO plans while meeting the rectal and urethral dose constraints. An IPSA class solution (CS-IPSA) was created from the mean MC-IPSA parameters. New plans were developed for each of these 40 patients using only CS-IPSA with no further optimization. Additionally, plans were created on an independent dataset of 30 different patients using only CS-IPSA with no further optimization. RESULTS Plans were optimized in about 30 minutes using GO, MC-IPSA took an average of 14.1 ± 6.5 minutes, and CS-IPSA optimizationwas < 1 minute. There was no significant difference (p > 0.05) among the optimization methods for all clinical goals over the 40 CS-IPSA source patients. There was no significant difference (p > 0.05) between the source and the independent datasets for all clinical goals when using CS-IPSA with no further optimization. CONCLUSION A prostate HDR IPSA class solution was developed and validated on a source and an independent dataset. The IPSA class solution yields plans comparable to custom manual IPSA and graphical optimization while saving considerable time.

SU‐E‐T‐453: Optimization of Dose Gradient for Gamma Knife Radiosurgery

PURPOSE The goals of stereotactic radiosurgery (SRS) are the ablation of target tissue and sparing of critical normal tissue. We develop tools to aid in the selection of collimation and prescription (Rx) isodose line to optimize the dose gradient for single isocenter intracranial stereotactic radiosurgery (SRS) with GammaKnife 4C utilizing the updated physics data in GammaPlan v10.1. METHODS Single isocenter intracranial SRS plans were created to treat the center of a solid water anthropomorphism head phantom for each GammaKnife collimator (4 mm, 8 mm, 14 mm, and 18 mm). The dose gradient, defined as the difference of effective radii of spheres equal to half and full Rx volumes, and Rx treatment volume was analyzed for isodoses from 99% to 20% of Rx. RESULTS The dosimetric data on Rx volume and dose gradient vs. Rx isodose for each collimator was compiled into an easy to read nomogram as well as plotted graphically. The 4, 8, 14, and 18 mm collimators have the sharpest dose gradient at the 64%, 70%, 76%, and 77% Rx isodose lines, respectively. This corresponds to treating 4.77 mm, 8.86 mm, 14.78 mm, and 18.77 mm diameter targets with dose gradients radii of 1.06 mm, 1.63 mm, 2.54 mm, and 3.17 mm, respectively. CONCLUSION We analyzed the dosimetric data for the most recent version of GammaPlan treatment planning software to develop tools that when applied clinically will aid in the selection of a collimator and Rx isodose line for optimal dose gradient and target coverage for single isocenter intracranial SRS with GammaKnife 4C.

SU-E-T-617: Dosimetric Comparison of Prone Breast Treatment on Tomotherapy and Conventional LINAC

Purpose: To report a comparative dosimetric analysis of prone breast treatments on Tomotherapy versus conventional LINAC. Methods: The static beam prone breast treatment with TomoDirect™ has been explored on Tomo treatment planning version 4.03 in comparison with conventional opposing beams in Philips Pinnacle version 9.0. Current LINAC based prone breast contours are imported into Tomo planning station and the same planning goals are established accordingly. Two laterally opposing beams from Tomotherapy are set with proper flashes to avoid any motion uncertainties. The common flash is opened for 4–5 leaves (each leaf is 6.25 mm) to properly compensate the breast margin. No forward segments need to be generated with TomoDirect, and the non‐uniform fluence sinogram pattern which will deliver homogeneous dose to cover the intended breast volume with a 3D compensation delivery nature. Results: Clinical dosimetry results indicate improved dosimetry coverage while minimizing the dose non‐uniformity inside the volume slices of PTV for Tomotherapy. With 97% coverage of the PTV prescription dose, segmented lateral breast treatments on the Pinnacle plan tends to have higher hot spots compared to TomoDirect plan (107.2% vs. 101.2%). Sparing of the right lung volume is also noticeable (1 cc of lung is 1425 cGy compared to 176 cGy in TomoDirect plan). Setup on the Tomotherapy unit is also easier since there will be no manual adjustment of the treatment couch. After MCVT, the imaged guidance matches the breast volume with automatic couch repositioning. Lung and heart sparing can be achieved via the inherent TomoDirect non‐uniform fluence delivery pattern with constant couch movement, which is similar to the 3D compensator approach in 3D breast treatment. Conclusions: With the Tomotherapy MVCT image guided approach and TomoDirect planning, prone breast can be easily positioned and delivered with better dose uniformity and minimized positional errors with reduced hot spots.

TU‐A‐BRA‐01: Non‐Coplanar Treatment of Hypofractionated Intracranial SRT with Tomotherapy

Purpose: To report on a newly developed positional device and dosimetric results in Tomotherapy intracranial hypofractionated Stereotactic Radiosurgery Treatment (SRT) with non‐coplanar beam characteristics. Methods: An in‐house developed positional device has been adapted to generate the non‐coplanar dosimetric effects in selected hypofractionated intracranial SRT patients. Three different head positions and CTdata sets of same patients were scanned to provide the basic panning information. Contrast is also provided to delineate CTV. Dose constraints for PTV and OARs were scaled down according to the total composite dose (30Gy is split into 12Gy, 12Gy and 6Gy with 5 treatment fractions). The ability to change the rotational angles of the head for each treatment fraction would reduce the volume of normal braintissue irradiated to lower clinically significant doses and increase the dose gradient surrounding the treatment area. Due to the TomoTherapy nature of helical co‐planar treatment, this methodology will generate a composite non‐coplanar delivery in order to reduce the low dose spread inside the PTV slices. Planning was performed with the three CTdata sets and the composite dosimetry was evaluated to prove the clinical efficacy. Results and Discussion: Clinical dosimetry results indicate improved dosimetry coverage while minimizing the scattered dose to the volume slices of PTV. Composite dosimetry indicated the 50% isodose volume minus target volume was reduced by 9.4% for a small tumor (0.7cm3). For a medium tumor (2.5cm3), the volume was reduced by 8.9% and for a large tumor (4.2cm3), the volume was reduced by 13.0%. All those reduction ha shown by manipulating the various head treatment positions, we can achieve lower dose sparing inside the volume of PTV slices. Conclusions: Our developed device and composite dosimetric results have shown the clinical benefits to improve the dose gradient and minimize the low dose region inside the PTV CT slices.

SU-FF-T-87: Dose Reconstruction of Intracranial Hypofractionated Helical Tomotherapy Treatments for Adaptive Planning

Purpose: To evaluate the dosimetry of delivered hypofractionated image guided helical tomotherapy treatments for adaptive planning. Method and Materials: Twelve patients with intracranial lesions received hypofractionated radiation treatments using helical tomotherapy. Image guidance MVCTs are merged with the planning kVCT images. The treatmentdeliveries then are calculated with a fine dose grid over the associated merged images. The summation dosimetry of the deliveredtreatments to the targets are analyzed using TomoTherapy adaptive planning software (TomoTherapy, Madison, WI) to determine if tuning the current plan to the patients daily treatment position would have been desirable. Results: On average, the difference between the planned treatments and deliveredtreatments in the coverage of the GTVs (n=33) by the prescription dose for all patients is 4.6%±13.1%. When selecting 99% of the prescription dose, the difference of mean GTV coverage between planned and deliveredtreatments reduces to 0.1%±0.4%. For the PTVs (n =33), the mean variation in coverage from the planned to deliveredtreatments by the prescription dose is 5.2%±10.0%. The average difference between planned versus deliveredtreatment coverage of the PTVs reduces to 0.5%±1.0% at 99% of the prescription dose.Conclusion: Due to the short course and high dose per delivery of hypofractionated radiation treatments, the importance of evaluating and if necessary adapting the planned treatment is pronounced for intracranial patients. As these IMRT plans create a sharp dose gradient there is notable incongruence between the planned and delivered coverage of the targets encompassed by the prescription dose. While changing the treatment plan to perform adaptive therapy seems to be the best solution, however, revision of the treated plans in this study would not be necessary as there is excellent agreement of target coverage between the planned and the deliveredimage guided helical tomotherapy treatments by the selected 99% of the prescription dose.

SU‐FF‐T‐664: Dose Grid Effects in Adaptive Planning of Helical TomoTherapy for Hypofractionated Treatments

Purpose: To evaluate potential dosimetry discrepancies generated by varying the dose grid resolution in adaptive planning of TomoTherapy for hypofractionated treatments.Method and Materials: Twelve patients with intracranial lesions treated with image guided helical TomoTherapy under a hypofractionated protocol are reviewed. Due to the short course and high dose per fraction, we use the adaptive planning tool to obtain the delivered dose distribution. The associated pretreatment MVCT and planning KVCT images are fused and the treatment doses are calculated comparing the fine dose grid (1.51×1.51×2mm3) and the normal grid (3.02×3.02×2mm3) settings. The summative dosimetry of the targets is analyzed to identify the effects of dose grid size on adaptive planning. Results: The mean difference in coverage of the GTVs by the prescription dose, calculated with fine versus normal dose grid, over all patients is 2.6%±10.5. When segregated by size the mean difference and standard deviation in GTV coverage for small lesions ( 7.5cc) lesions at 0.6%±0.6 and 1.4%±2.5, respectively. For all the patients the mean variation between the fine and normal grids in the calculated coverage of the PTVs by the prescription dose is 16.3%±17.4. For small PTVs alone ( 15cc). Conclusion: While performing critical adaptive planning evaluation for intracranial patients treated with TomoTherapy, influence of the dose grid on the summation dosimetry must be considered. In our study, there is an appreciable difference in calculated target coverage amongst different dose grid resolutions, especially for small targets treated under hypofractionated protocols. Consequently, the use of fine dose grid is necessary if adaptive planning is performed for assessing positioning errors.

SU‐GG‐J‐55: Comparison of Megavoltage Imaging Modalities and Spatial Effects for Dose Calculations.

Purpose: To investigate spatial effects of differing physical densities in megavoltage imaging with cone‐beam computed tomography (MVCB) versus helical tomotherapy (MVCT) for dose calculation. Method and Materials:Image value‐to‐density tables (IVDT) were created for MVCB and MVCT against standard kilovoltge computed tomography (KVCT) using a solid‐water phantom and varying tissue equivalent inserts. The spatial independence of the different physical densities of the image was tested by generating two sets of IVDT for each modality: one with the inserts in the peripheral locations in the phantom and another with the inserts in the inner locations. Scans of an anthropomorphic head phantom for each modality were imported into a treatment planning system; contours of anatomy were created and a seven field IMRT plan was developed to simulate treatment to a target in the brain. The dose was then calculated for each imaging modality scan corrected with its given IVDT. Results: The IVDT sets for KVCT were nearly identical to each other; the IVDT sets for MVCT were practically equal also. Such similarity was confirmed by comparing the dose volume histograms of treatment plans; average coverage of the target was 95.5±0.2%. However, the IVDT sets for MVCB exhibited an appreciable difference as the central region of the images had a lower value inaccurately reducing the density. This was reflected in the MVCB treatment plans as the prescription covered 17.75% of the target using IVDT with tissue equivalent inserts in the inner locations. This coverage was increased to 98.65% using the outer locations. Conclusion: KVCT and MVCT modalities provide images suitable for dose calculation of differing physical densities that is independent of their spatial location. However, the artifacts introduced by MVCB imaging can lead to notable differences in image values and its corresponding densities that render current MVCB unsuitable for dose calculation.