Search

In the electron beam radiation therapy, customized blocks are mostly used to shape treatment fields to generate conformal doses. The goal of this study is to investigate quantitatively dosimetric uncertainties associated with heterogeneities, detectors used in the measurement of the beam data commissioning, and modeling of the interactions of high energy electrons with tissue. These uncertainties were investigated both by measurements with different detectors and calculations using electron Monte Carlo algorithm (eMC) in the Eclipse treatment planning system. Dose distributions for different field sizes were calculated using eMC and measured with a multiple‐diode‐array detector (MapCheck2) for cone sizes ranging from 6 to 25 cm. The dose distributions were calculated using the CT images of the MapCheck2 and water‐equivalent phantoms. In the umbra region (

The goal of this project is to investigate quantitatively the performance of different deformable image registration algorithms (DIR) with helical (HCT), axial (ACT), and cone‐beam CT (CBCT). The variations in the CT‐number values and lengths of well‐known targets moving with controlled motion were evaluated. Four DIR algorithms: Demons, Fast‐Demons, Horn‐Schunck and Lucas‐Kanade were used to register intramodality CT images of a mobile phantom scanned with different imaging techniques. The phantom had three water‐equivalent targets inserted in a low‐density foam with different lengths (10–40 mm) and moved with adjustable motion amplitudes (0–20 mm) and frequencies (0–0.5 Hz). The variations in the CT‐number level, volumes and shapes of these targets were measured from the spread‐out of the CT‐number distributions. In CBCT, most of the DIR algorithms were able to produce the actual lengths of the mobile targets; however, the CT‐number values obtained from the DIR algorithms deviated from the actual CT‐number of the targets. In HCT, the DIR algorithms were successful in deforming the images of the mobile targets to the images of the stationary targets producing the CT‐number values and lengths of the targets for motion amplitudes

The purpose of this study was to assess quantitatively elongation of mobile targets in cone‐beam CT (CBCT) imaging by measurement and modeling. A mathematical model was derived that predicts the measured lengths of mobile targets and its dependence on target size and motion patterns in CBCT imaging. Three tissue‐equivalent targets of differing sizes were inserted in an artificial thorax phantom to simulate lung lesions. Respiratory motion was mimicked with a mobile phantom that moves in one‐dimension along the superior‐inferior direction at a respiration frequency of 0.24 Hz for eight different amplitudes in the range 0‐40 mm. A mathematical model was derived to quantify the variations in target lengths and its dependence on phantom motion parameters in CBCT. Predictions of the model were verified by measurement of the lengths of mobile targets in CBCT images. The model predicts that target lengths increased linearly with increase in speed and amplitude of phantom motion in CBCT. The measured lengths of mobile targets imaged with CBCT agreed with the calculated lengths within half‐slice thickness spatial resolution. The maximal length of a mobile target was independent of the frequency and phase of motion. Elongation of mobile targets was similar in halffan and full‐fan CBCT for similar motion patterns, as long as the targets remained within the imaging view. Mobile targets elongated linearly with phantom speed and motion amplitude in CBCT imaging. The model introduced in this work assessed quantitatively the variation in target lengths induced by motion, which may be a useful tool to consider elongations of mobile targets in CBCT applications in diagnostic imaging and radiotherapy. PACS number: 87.57.qp

Introduction We aim to quantify the variations in the gross tumour volume (GTV) during a course of stereotactic body radiotherapy (SBRT) and determine its impact on dosimetric coverage of the GTV. Methods The GTVs and dose coverage for 14 patients with 16 primary non-small-cell lung tumours treated with SBRT were investigated. Initial GTVs were calculated from treatment planning CT scans. The prescribed doses ranged from 48 to 60 Gy in three to five fractions. Before each treatment, patients underwent a CBCT scan. For each CBCT scan, the GTV and the dose received by the GTV were determined and followed during the course of therapy. Results There was considerable variation in the measured GTVs during the course of therapy. Increases of up to 63.3% of volume measured by initial CBCT were detected during the first few fractions, after which GTV tended to decrease. Dose coverage (V95) for any given fraction deviated no more than 5% from optimised coverage obtained in the initial treatment plan. In the long term, all patients with follow-up scans demonstrated tumour shrinkage with no radiographic evidence of tumour recurrence. Conclusion GTV, as evaluated in this study, demonstrates an initial increase in volume followed by a subsequent decrease. This volume change needs to be considered in the design of treatment plans and assignment of treatment margins.

Two recently introduced GafChromic film models, HS and XR-T, have been developed as more sensitive and uniform alternatives to GafChromic MD-55-2 film. The HS model has been specifically designed for measurement of absorbed dose in high-energy photon beams (above 1 MeV), while the XR-T model has been introduced for dose measurements of low energy (0.1 MeV) photons. The goal of this study is to compare the sensitometric curves and estimated dosimetric uncertainties associated with seven different GafChromic film dosimetry systems for the two new film models. The densitometers tested are: LKB Pharmacia UltroScan XL, Molecular Dynamics Personal Densitometer, Nuclear Associates Radiochromic Densitometer Model 37-443, Photoelectron Corporation CMR-604, Laser Pro 16, Vidar VXR-16, and AGFA Arcus II document scanner. Pieces of film were exposed to different doses in a dose range from 0.5 to 50 Gy using 6 MV photon beam. Functional forms for dose vs net optical density have been determined for each of the GafChromic film-dosimetry systems used in this comparison. Two sources of uncertainties in dose measurements, governed by the experimental measurement and calibration curve fit procedure, have been compared for the densitometers used. Among the densitometers tested, it is found that for the HS film type the uncertainty caused by the experimental measurement varies from 1% to 3% while the calibration fit uncertainty ranges from 2% to 4% for doses above 5 Gy. Corresponding uncertainties for XR-T film model are somewhat higher and range from 1% to 5% for experimental and from 2% to 7% for the fit uncertainty estimates. Notwithstanding the significant variations in sensitivity, the studied densitometers exhibit very similar precision for GafChromic film based dose measurements above 5 Gy.

Purpose: To investigate quantitatively the accuracy of dose distributions for the Ir-192 high-dose-rate (HDR) brachytherapy source calculated by the Brachytherapy-Planning system (BPS) and measured using a multiple-array-diode-detector in a heterogeneous medium. Methods: A two-dimensional diode-array-detector system (MapCheck2) was scanned with a catheter and the CT-images were loaded into the Varian-Brachytherapy-Planning which uses TG-43-formalism for dose calculation. Treatment plans were calculated for different combinations of one dwell-position and varying irradiation times and different-dwell positions and fixed irradiation time with the source placed 12mm from the diode-array plane. The calculated dose distributions were compared to the measured doses with MapCheck2 delivered by an Ir-192-source from a Nucletron-Microselectron-V2-remote-after-loader. The linearity of MapCheck2 was tested for a range of dwell-times (2–600 seconds). The angular effect was tested with 30 seconds irradiation delivered to the central-diode and then moving the source away in increments of 10mm. Results: Large differences were found between calculated and measured dose distributions. These differences are mainly due to absence of heterogeneity in the dose calculation and diode-artifacts in the measurements. The dose differences between measured and calculated due to heterogeneity ranged from 5%–12% depending on the position of the source relative to the diodes in MapCheck2 and different heterogeneities in the beam path. The linearity test of the diode-detector showed 3.98%, 2.61%, and 2.27% over-response at short irradiation times of 2, 5, and 10 seconds, respectively, and within 2% for 20 to 600 seconds (p-value=0.05) which depends strongly on MapCheck2 noise. The angular dependency was more pronounced at acute angles ranging up to 34% at 5.7 degrees. Conclusion: Large deviations between measured and calculated dose distributions for HDR-brachytherapy with Ir-192 may be improved when considering medium heterogeneity and dose-artifact of the diodes. This study demonstrates that multiple-array-diode-detectors provide practical and accurate dosimeter to verify doses delivered from the brachytherapy Ir-192-source.

Purpose: To evaluate quantitatively dose distributions from helical, axial and cone-beam CT clinical imaging techniques by measurement using a two-dimensional (2D) diode-array detector. Methods: 2D-dose distributions from selected clinical protocols used for axial, helical and cone-beam CT imaging were measured using a diode-array detector (MapCheck2). The MapCheck2 is composed from solid state diode detectors that are arranged in horizontal and vertical lines with a spacing of 10 mm. A GE-Light-Speed CT-simulator was used to acquire axial and helical CT images and a kV on-board-imager integrated with a Varian TrueBeam-STx machine was used to acquire cone-beam CT (CBCT) images. Results: The dose distributions from axial, helical and cone-beam CT were non-uniform over the region-of-interest with strong spatial and angular dependence. In axial CT, a large dose gradient was measured that decreased from lateral sides to the middle of the phantom due to large superficial dose at the side of the phantom in comparison with larger beam attenuation at the center. The dose decreased at the superior and inferior regions in comparison to the center of the phantom in axial CT. An asymmetry was found between the right-left or superior-inferior sides of the phantom which possibly to angular dependence in the dosemore » distributions. The dose level and distribution varied from one imaging technique into another. For the pelvis technique, axial CT deposited a mean dose of 3.67 cGy, helical CT deposited a mean dose of 1.59 cGy, and CBCT deposited a mean dose of 1.62 cGy. Conclusions: MapCheck2 provides a robust tool to measure directly 2D-dose distributions for CT imaging with high spatial resolution detectors in comparison with ionization chamber that provides a single point measurement or an average dose to the phantom. The dose distributions measured with MapCheck2 consider medium heterogeneity and can represent specific patient dose.« less

Purpose: To compare the impact of MLC width on the dose conformity, doses received by the normal brain tissue, and other critical structures for single isocenter (SI) and multiple isocenter (MI) stereotactic IMRT (SRT) treatments for patients with at least four brain tumors. Methods: Six patients (three with seven and three with four lesions) planned in BrainLab iPlan were evaluated using Varian HD MLC (2.5mm leaf-width) and Millennium-120 MLC (5mm leaf-width). Prescribed doses in all plans were 25 Gy delivered in five fractions normalized to 98% dose covering 98% of the target volume. Results: All plans were judged clinically acceptable. The average Paddick conformity index for 2.5mm vs 5mm SI and MI plans were (0.48±0.22 vs. 0.38±0.21, p

Purpose: To evaluate the performance of a two-dimensional (2D) array-diode- detector for geometric and dosimetric quality assurance (QA) tests of high-dose-rate (HDR) brachytherapy with an Ir-192-source. Methods: A phantom setup was designed that encapsulated a two-dimensional (2D) array-diode-detector (MapCheck2) and a catheter for the HDR brachytherapy Ir-192 source. This setup was used to perform both geometric and dosimetric quality assurance for the HDR-Ir192 source. The geometric tests included: (a) measurement of the position of the source and (b) spacing between different dwell positions. The dosimteric tests include: (a) linearity of output with time, (b) end effect and (c) relative dose verification. The 2D-dose distribution measured with MapCheck2 was used to perform the previous tests. The results of MapCheck2 were compared with the corresponding quality assurance testes performed with Gafchromic-film and well-ionization-chamber. Results: The position of the source and the spacing between different dwell-positions were reproducible within 1 mm accuracy by measuring the position of maximal dose using MapCheck2 in contrast to the film which showed a blurred image of the dwell positions due to limited film sensitivity to irradiation. The linearity of the dose with dwell times measured from MapCheck2 was superior to the linearity measured with ionization chamber due to higher signal-to-noise ratio of the diode readings. MapCheck2 provided more accurate measurement of the end effect with uncertainty < 1.5% in comparison with the ionization chamber uncertainty of 3%. Although MapCheck2 did not provide absolute calibration dosimeter for the activity of the source, it provided accurate tool for relative dose verification in HDR-brachytherapy. Conclusion: The 2D-array-diode-detector provides a practical, compact and accurate tool to perform quality assurance for HDR-brachytherapy with an Ir-192 source. The diodes in MapCheck2 have high radiation sensitivity and linearity that is superior to Gafchromic-films and ionization chamber used for geometric and dosimetric QA in HDR-brachytherapy, respectively.

Purpose: To quantitatively investigate the surface dose deposited in patients imaged with a kV on-board-imager mounted on a radiotherapy machine using different clinical imaging techniques and filters. Methods: A high sensitivity photon diode is used to measure the surface dose on central-axis and at an off-axis-point which is mounted on the top of a phantom setup. The dose is measured for different imaging techniques that include: AP-Pelvis, AP-Head, AP-Abdomen, AP-Thorax, and Extremity. The dose measurements from these imaging techniques are combined with various filtering techniques that include: no-filter (open-field), half-fan bowtie (HF), full-fan bowtie (FF) and Cu-plate filters. The relative surface dose for different imaging and filtering techniques is evaluated quantiatively by the ratio of the dose relative to the Cu-plate filter. Results: The lowest surface dose is deposited with the Cu-plate filter. The highest surface dose deposited results from open fields without filter and it is nearly a factor of 8–30 larger than the corresponding imaging technique with the Cu-plate filter. The AP-Abdomen technique delivers the largest surface dose that is nearly 2.7 times larger than the AP-Head technique. The smallest surface dose is obtained from the Extremity imaging technique. Imaging with bowtie filters decreases the surface dose by nearly 33% in comparison with the open field. The surface doses deposited with the HF or FF-bowtie filters are within few percentages. Image-quality of the radiographic images obtained from the different filtering techniques is similar because the Cu-plate eliminates low-energy photons. The HF- and FF-bowtie filters generate intensity-gradients in the radiographs which affects image-quality in the different imaging technique. Conclusion: Surface dose from kV-imaging decreases significantly with the Cu-plate and bowtie-filters compared to imaging without filters using open-field beams. The use of Cu-plate filter does not affect image-quality and may be used as the default in the different imaging techniques.

Purpose: To investigate quantitatively positioning and dosimetric uncertainties due to 4D-CT intra-phase motion in the internal-target-volume (ITV) associated with radiation therapy using respiratory-gating for patients setup with image-guidance-radiation-therapy (IGRT) using free-breathing or average-phase CT-images. Methods: A lung phantom with an embedded tissue-equivalent target is imaged with CT while it is stationary and moving. Four-sets of structures are outlined: (a) the actual target on CT-images of the stationary-target, (b) ITV on CT-images for the free-moving phantom, (c) ITV's from the ten different phases (10–100%) and (d) ITV on the CT-images generated from combining 3 phases: 40%–50%–60%. The variations in volume, length and center-position of the ITV's and their effects on dosimetry during dose delivery for patients setup with image-guidance are investigated. Results: Intra-phase motion due to breathing affects the volume, center position and length of the ITVs from different respiratory-phases. The ITV's vary by about 10% from one phase to another. The largest ITV is measured on the free breathing CT images and the smallest is on the stationary CT-images. The ITV lengths vary by about 4mm where it may shrink or elongated depending on the motion-phase. The center position of the ITV varies between the different motion-phases which shifts upto 10mm from the stationary-position which is nearly equal to motion-amplitude. This causes systematic shifts during dose delivery with beam gating using certain phases (40%–50%–60%) for patients setup with IGRT using free-breathing or average-phase CT-images. The dose coverage of the ITV depends on the margins used for treatment-planning-volume where margins larger than the motion-amplitudes are needed to ensure dose coverage of the ITV. Conclusion: Volume, length, and center position of the ITV's change between the different motion phases. Large systematic shifts are induced by respiratory-gating with ITVs on certain phases when patients are setup with IGRT using free-breathing or average-phase CT-images.

Purpose: To develop accurate three-dimensional (3D) empirical dose calculation model for kV diagnostic beams for different radiographic and CT imaging techniques. Methods: Dose was modeled using photon attenuation measured using depth dose (DD), scatter radiation of the source and medium, and off-axis ratio (OAR) profiles. Measurements were performed using single-diode in water and a diode-array detector (MapCHECK2) with kV on-board imagers (OBI) integrated with Varian TrueBeam and Trilogy linacs. The dose parameters were measured for three energies: 80, 100, and 125 kVp with and without bowtie filters using field sizes 1×1–40×40 cm2 and depths 0–20 cm in water tank. Results: The measured DD decreased with depth in water because of photon attenuation, while it increased with field size due to increased scatter radiation from medium. DD curves varied with energy and filters where they increased with higher energies and beam hardening from half-fan and full-fan bowtie filters. Scatter radiation factors increased with field sizes and higher energies. The OAR was with 3% for beam profiles within the flat dose regions. The heal effect of this kV OBI system was within 6% from the central axis value at different depths. The presence of bowtie filters attenuated measured dose off-axis by asmore » much as 80% at the edges of large beams. The model dose predictions were verified with measured doses using single point diode and ionization chamber or two-dimensional diode-array detectors inserted in solid water phantoms. Conclusion: This empirical model enables fast and accurate 3D dose calculation in water within 5% in regions with near charge-particle equilibrium conditions outside buildup region and penumbra. It considers accurately scatter radiation contribution in water which is superior to air-kerma or CTDI dose measurements used usually in dose calculation for diagnostic imaging beams. Considering heterogeneity corrections in this model will enable patient specific dose calculation.« less

Purpose: To evaluate patient setup accuracy and quantify individual and cumulative positioning uncertainties associated with different hardware and software components of the stereotactic radiotherapy (SRS/SRT) with the frameless-6D-ExacTrac system. Methods: A statistical model was used to evaluate positioning uncertainties of the different components of SRS/SRT treatment with the BrainLAB 6D-ExacTrac system using the positioning shifts of 35 patients having cranial lesions (49 total lesions treated in 1, 3, 5 fractions). All these patients were immobilized with rigid head-and-neck masks, simulated with BrainLAB-localizer and planned with iPlan treatment planning system. Infrared imaging (IR) was used initially to setup patients. Then, stereoscopic x-ray images (XC) were acquired and registered to corresponding digitally-reconstructed-radiographs using bony-anatomy matching to calculate 6D-translational and rotational shifts. When the shifts were within tolerance (0.7mm and 1°), treatment was initiated. Otherwise corrections were applied and additional x-rays were acquired (XV) to verify that patient position was within tolerance. Results: The uncertainties from the mask, localizer, IR-frame, x-ray imaging, MV and kV isocentricity were quantified individually. Mask uncertainty (Translational: Lateral, Longitudinal, Vertical; Rotational: Pitch, Roll, Yaw) was the largest and varied with patients in the range (−1.05−1.50mm, −5.06–3.57mm, −5.51−3.49mm; −1.40−2.40°, −1.24−1.74°, and −2.43−1.90°) obtained from mean of XC shifts for each patient. Setup uncertainty in IR positioning (0.88,2.12,1.40mm, and 0.64,0.83,0.96°) was extracted from standard-deviation of XC. Systematic uncertainties of the localizer (−0.03,−0.01,0.03mm, and −0.03,0.00,−0.01°) and frame (0.18,0.25,−1.27mm,−0.32,0.18, and 0.47°) were extracted from means of all XV setups and mean of all XC distributions, respectively. Uncertainties in isocentricity of the MV radiotherapy machine were (0.27,0.24,0.34mm) and kV-imager (0.15,−0.4,0.21mm). Conclusion: A statistical model was developed to evaluate the individual and cumulative systematic and random uncertainties induced by the different hardware and software components of the 6D-ExacTrac-system. The immobilization mask was associated with the largest positioning uncertainty.

Purpose: Clinically relevant dose errors of small field brainIMRT were evaluated using planar per‐beam dosimetry and planned dose perturbation (PDP; used in the Sun Nuclear 3DVH+MapCHEKC2) algorithm. Methods: Ten BrainLAB IMRT plans were retrospectively selected. PTVs ranged from 0.5 to 29.0 cc and most of the field sizes were smaller than 4×4 cm2 (beamlets were much smaller). The Varian TrueBEAM STx with high‐definition MLC (2.5 mm width at isocenter) was employed to deliver treatment plans. The conventional per‐beam IMRT QA was performed using the 2D diode array (diode detector size of 0.7×0.7 mm2). Two shifted QA measurements were merged for each plan to achieve a grid spacing of 5 mm. The 3DVH reconstructed a ‘true’ DVH of plans using the measured errors back‐projected on to 3D patient dose calculation. Dose coverage in PTVs and critical structures was compared between treatment planning and actual delivery. 2D (conventional planar QA) and 3D (PDP‐reconstructed dose) gamma analyses were also performed. Results: In DVH‐based analysis using 3DVH, the percent differences of maximum dose between planning and delivery for PTV were −0.4 to 20.2% (p‐value: 0.02) and for normal tissue −5.67 to 1.32% (brainstem), −2.50 to 2.51% (eye), and −4.9 to 4.9% (optic nerve). The D95 percent differences for PTV were −3.2 to −0.56% (p‐value: 0.0001). Average pass rates (1 standard deviation) of the gamma test with 3%‐3 mm criteria were 95.7 (4.9)% (2D) and 97.6 (3.4)% (3D), respectively. The correlation coefficients between 2D/3D gamma tests and the change in PTV coverage were 0.76/0.68 (Dmax of PTV) and −0.86/−0.96 (D95 of PTV), respectively. Conclusions: There was a strong correlation between the gamma analysis and the predicted PTV coverage in small field dosimetry. Using 3D reconstructed dose maps, statistically significant dose difference between planning and delivery was observed in PTVs for the small field IMRT.

Purpose: To investigate the doses received by the hippocampus and normal brain tissue during a course of stereotactic radiotherapy utilizing a single isocenter (SI) versus multiple isocenter (MI) in patients with multiple intracranial metastases. Methods: Seven patients imaged with MRI including SPGR sequence and diagnosed with 2–3 brain metastases were included in this retrospective study. Two sets of stereotactic IMRT treatment plans, (MI vs SI), were generated. The hippocampus was contoured on SPGR sequences and doses received by the hippocampus and whole brain were calculated. The prescribed dose was 25Gy in 5 fractions. The two groups were compared using t-test analysis. Results: There were 17 lesions in 7 patients. The median tumor, right hippocampus, left hippocampus and brain volumes were: 3.37cc, 2.56cc, 3.28cc, and 1417cc respectively. In comparing the two treatment plans, there was no difference in the PTV coverage except in the tail of the DVH curve. All tumors had V95 > 99.5%. The only statistically significant parameter was the V100 (72% vs 45%, p=0.002, favoring MI). All other evaluated parameters including the V95 and V98 did not reveal any statistically significant differences. None of the evaluated dosimetric parameters for the hippocampus (V100, V80, V60, V40, V20, V10, D100, D90, D70, D50, D30, D10) revealed any statistically significant differences (all p-values > 0.31) between MI and SI plans. The total brain dose was slightly higher in the SI plans, especially in the lower dose regions, although this difference was not statistically significant. Utilizing brain-sub-PTV volumes did not change these results. Conclusion: The use of SI treatment planning for patients with up to 3 brain metastases produces similar PTV coverage and similar normal tissue doses to the hippocampus and the brain compared to MI plans. SI treatment planning should be considered in patients with multiple brain metastases undergoing stereotactic treatment.

Purpose: To evaluate setup accuracy and quantify individual systematic and random errors for the various hardware and software components of the frameless 6D-BrainLAB ExacTrac system. Methods: 35 patients with cranial lesions, some with multiple isocenters (50 total lesions treated in 1, 3, 5 fractions), were investigated. All patients were simulated with a rigid head-and-neck mask and the BrainLAB localizer. CT images were transferred to the IPLAN treatment planning system where optimized plans were generated using stereotactic reference frame based on the localizer. The patients were setup initially with infrared (IR) positioning ExacTrac system. Stereoscopic X-ray images (XC: X-ray Correction) were registered to their corresponding digitally-reconstructed-radiographs, based on bony anatomy matching, to calculate 6D-translational and rotational (Lateral, Longitudinal, Vertical, Pitch, Roll, Yaw) shifts. XC combines systematic errors of the mask, localizer, image registration, frame, and IR. If shifts were below tolerance (0.7 mm translational and 1 degree rotational), treatment was initiated; otherwise corrections were applied and additional X-rays were acquired to verify patient position (XV: X-ray Verification). Statistical analysis was used to extract systematic and random errors of the different components of the 6D-ExacTrac system and evaluate the cumulative setup accuracy. Results: Mask systematic errors (translational; rotational) were the largest and varied from one patient to another in the range (−15 to 4mm; −2.5 to 2.5degree) obtained from mean of XC for each patient. Setup uncertainty in IR positioning (0.97,2.47,1.62mm;0.65,0.84,0.96degree) was extracted from standard-deviation of XC. Combined systematic errors of the frame and localizer (0.32,−0.42,−1.21mm; −0.27,0.34,0.26degree) was extracted from mean of means of XC distributions. Final patient setup uncertainty was obtained from the standard deviations of XV (0.57,0.77,0.67mm,0.39,0.35,0.30degree). Conclusion: Statistical analysis was used to calculate cumulative and individual systematic errors from the different hardware and software components of the 6D-ExacTrac-system. Patients were treated with cumulative errors (

Purpose: To assess the impacts that multileaf collimator (MLC) leaf width has on the dose conformity and normal brain tissue doses of single and multiple isocenter stereotactic IMRT (SRT) plans for multiple intracranial tumors. Methods: Fourteen patients with 2–3 targets were studied retrospectively. Patients treated with multiple isocenter treatment plans using 9 to 12 non-coplanar beams per lesion underwent repeat planning using single isocenter and 10 to 12 non-coplanar beams with 2.5mm, 3mm and 5mm MLC leaf widths. Brainlab iPlan treatment planning system for delivery with the 2.5mm MLC served as reference. Identical contour sets and dose-volume constraints were applied. The prescribed dose to each target was 25 Gy to be delivered over 5 fractions with a minimum of 99% dose to cover ≥ 95% of the target volume. Results: The lesions and normal brains ranged in size from 0.11 to 51.67cc (median, 2.75cc) and 1090 to 1641cc (median, 1401cc), respectively. The Paddick conformity index for single and multiple isocenter (2.5mm vs. 3mm and 5mm MLCs) was (0.79±0.08 vs. 0.79±0.07 and 0.77±0.08) and (0.79±0.09 vs. 0.77±0.09 and 0.76±0.08), respectively. The average normal brain volumes receiving 15 Gy for single and multiple isocenter (2.5mm vs. 3mm and 5mm MLCs) were (3.65% vs. 3.95% and 4.09%) and (2.89% vs. 2.91% and 2.92%), respectively. Conclusion: The average dose conformity observed for the different leaf width for single and multiple isocenter plans were similar, throughout. However, the average normal brain volumes receiving 2.5 to 15 Gy were consistently lower for the 2.5mm MLC leaf width, especially for single isocenter plans. The clinical consequences of these integral normal brain tissue doses are still unknown, but employing the use of the 2.5mm MLC option is desirable at sparing normal brain tissue for both single and multiple isocenter cases.

Purpose: To compare quality of single and multiple isocenter stereotactic IMRT treatment plans for multiple intracranial targets. Methods: Five patients (two with three and three with two brain metastatic tumors) treated with multiple isocenters using 9 to 12 non‐coplanar beams per lesion were replanned for single isocenter 10 to 12 non‐coplanar beams. Identical contour sets and dose‐volume constraints were applied. Plans were developed using BrainLab iPlan system for delivery with Varian TrueBeam STx machine. The prescribed dose to each target was 25 Gy in five fractions and minimum 95% dose was required to cover > 95% of target volume. Results: All plans satisfied the dose‐volume constraints, with no significant differences in dose conformity between single and multiple isocenter plans. The mean Paddick conformity index (CI) for two target single and multiple isocenter plans was 0.76 and 0.75 (p = 0.93), and for three target single and multiple isocenter plans was 0.73 and 0.70 (p = 0.37), respectively. Normal brain tissue doses were higher for single isocenter plans compared to multiple isocenter plans. The average normal brain volumes receiving 15 and 10 Gy for two target plans (single vs. multiple isocenter) were: (4.8% vs. 2.1%) and (8.9% vs. 4.5%), respectively. For three target plans these volumes were: (2.0% vs. 1.5%) and (4.6% vs. 3.4%). The estimated total treatment time (patient set‐up, imaging‐verification, beam‐on time at 600 MU/min, etc.) were approximately 1.75 and 2.5 times lower with two and three target single isocenter delivery, respectively. Conclusion: Single isocenter plans were capable of producing quality treatment plans similar to multiple isocenter plans. Treatment delivery times were much higher and the normal brain tissue doses were only slightly lower at 10 and 15 Gy dose levels for multiple isocenter plans compared to single isocenter plans. Clinical consequences of these normal tissue doses are however unknown.

Purpose: To investigate the effects of optical density (OD) sensitivity and growth with time post‐irradiation on intensity‐modulated radiation therapy(IMRT)dosimetry with gafchromic films (EBT1 and EBT2). Method and Materials:EBT1 and EBT2 films were used to perform IMRTquality assurance, which were exposed to IMRT plans created in treatment planning system using 6 MV photon beam from a Trilogy Varian linac.Calibrationfilms were exposed to graded doses from 78–399cGy in order to convert OD to dose. The films were scanned using a flat panel scanner several times starting from 4 to 480 hours post‐irradiation. Gray scale and red colors for both EBT1 and EBT2 films were used to evaluate film sensitivity on IMRT analysis. The IMRTfilms were analyzed with calibrationfilms that were scanned synchronously and at different times post‐irradiation to quantify the effect of OD growth.Dose difference and gamma analysis of the dose distributions measured by film and calculated by treatment planning system were used to evaluate the effects of OD growth and sensitivity. Result: OD variations show that EBT1 grows slower, however, reaching higher values (14%) than EBT2 (13%) after long time post‐irradiation. The gray color values grow higher (13–14%) than red values (6%) for both EBT1&2 films after 480 hours. The results of dose difference and gamma parameter with more than 85% and 97% passing rates, respectively, using acceptance criteria (5%, 5mm) were better for films analyzed with synchronized calibration and red color values than non‐synchronized calibration and grayscale. Conclusion: The measurement of dose distributions agrees well with those calculated by treatment planning for synchronized IMRT and calibrationfilms where variation due to OD growth is minimized. Red color values provide better gamma for IMRT analysis with EBT1&2 films because of lower OD growth and high sensitivity than gray color values.

Purpose: To investigate correlation of standard‐uptake values (SUV) from post‐treatment PETimaging representing clinical complications such as muscular inflammation with dose delivered in non‐small cell lungcancer patients treated with large hypo‐fractions using stereotactic body radiation therapy(SBRT). Method and Materials: Clinical complications represented as chronic inflammation and rib fractures in ten non‐small cell lungcancer patients treated with large hypo‐fractionated stereotactic body radiation therapy were investigated. SBRT was delivered in three fractions with each fraction 20 Gy or five fractions with each 12 Gy. These patients were scanned with PETimaging at three and seven months after treatment for follow up to evaluate tumor control and clinical complications post‐ radiotherapy. The dose distributions calculated by treatment planning and delivered were extracted and compared with SUV distributions obtained from PETimaging. The correlation between dose deposited in the chest wall and clinical complications represented by higher SUV from PETimaging were quantified. Results: The dose distribution calculated by treatment planning from the different beams overlaid on a CT axial image for a lung patient correlates strongly with SUV obtained from PETimaging. The enhanced SUV represents chronic inflammation of the chest wall tissue that has received high doses by the individual entering beams. The profiles from the regions in the chest wall that have received high dose deposition correspond to enhanced SUV in anatomically matching regions from PETimages. Significant proportion of pixels from dose distribution correlates linearly with corresponding pixels that have high SUV‐dose in the chest wall. Conclusions: Large doses deposited in normal tissue from SBRT in treatment of non‐small cell lungtumor correlate well with the SUV values measured by PETimages. This correlation may be useful in predicting clinical complications by measurement of the delivered dose and SUV values associated with SBRT.

Purpose: To compare and assess accuracy of the doses calculated using the BrainLAB pencil beam (PB) and Monte Carlo(MC) algorithms for lung, prostate, brain, head and neck and paraspinal tumors.Methods and Material: Dose was calculated using PB convolution and MC algorithms in the IPLAN treatment planning system from BrainLAB for 5 lungs, 3 brains, 5 prostates, 2 head and neck and 2 paraspinal tumors. Dose was calculated using a combination of three‐dimensional conformai and IMRT plans. The leaf sequences from IMRT plans or beam shapes from conformai plan and monitor units (MUs) calculated by the PB were used to calculate dose with MC.Results: The DVH's calculated by PB and MC in the brain, prostate, paraspinal and head and neck are in good agreement within 5%. However, the DVH's of the lung patients have large discrepancies and while PB shows good coverage, MC shows lack of dose coverage. For maintaining similar dose coverage, higher MUs up to 30% are needed for MC compared to that predicted by PB calculations. Despite large discrepancies in DVH coverage of the PTV between PB and MC, the point dose at isocenter for lung calculated by both algorithms were within 5%. Further, the dose measured at isocenter using ionization chamber agrees well with pencil beam within 5% for several selected plans. Conclusions: The calculated dose by MC and PB agrees within 5% for prostate, brain, head and neck and paraspinal tumors. However, considerable discrepancies up to 30% are observed in dose‐volume coverage between MC and PB in lungtumors. Point dose measurements at isocenter are not representative of the discrepancies in dose coverage between PB and MC and verification with two‐ and three‐dimensional dose measurements are required.

Purpose: To evaluate setup accuracy of the BrainLAB stereotactic radiotherapy head immobilization mask using orthogonal projections from on‐board imaging.Materials and Methods: Eight patients were simulated and treated with the BrainLAB stereotactic head immobilization masks and CT‐localizers for brain lesions using single and hypo‐fractions. Orthogonal pairs of kV projections were acquired using Varian Trilogy's on‐board imager. The kV projections were then registered with digitally‐reconstructed‐radiographs (DRR) obtained from treatment planning. Shifts between the kV images and reference DRR's were calculated: anterior‐posterior (A‐P), left‐right (L‐R) and superior‐inferior (S‐I). If the shifts were > 2 mm in any direction, the patient was reset within the immobilization mask until satisfying accuracy has been achieved. Results: Shifts as large as 5 mm in the A‐P, L‐R and S‐I directions between kV projections and DRR's were found. These shifts represent offsets between the treatment and simulation setup using immobilization mask. The mean offsets of 0.1 mm, 0.7 mm, and −1.6 mm represent systematic errors in the A‐P, L‐R and S‐I directions, respectively, using the BrainLAB localizer. The mean of the radial shifts is about 1.7 mm. The standard deviations of shifts 2.2 mm, 2.0 mm, and 2.6 mm represent patient setup accuracy with the BrainLAB mask in A‐P, L‐R and S‐I directions. Conclusions: The BrainLAB mask provides less invasive immobilization system than the head ring. Relying on this system for patient setup might be associated with errors as large as 5 mm. Image‐guidance with the kV on‐board imager provides shifts that are required to setup patient prior to or during patient treatment. The patient may relax or move during treatment, and thus uncontrolled and undetected setup errors may be produced with patients that are not well immobilized. The combination of stereotactic immobilization and image guidance will achieve more controlled and accurate patient setup within 2 mm.

Purpose: To evaluate setup accuracy of the BrainLAB stereotactic radiotherapy head immobilization mask using orthogonal projections from on‐board imaging.Materials and Methods: Eight patients were simulated and treated with the BrainLAB stereotactic head immobilization masks and CT‐localizers for brain lesions using single and hypo‐fractions. Orthogonal pairs of kV projections were acquired using Varian Trilogy's on‐board imager. The kV projections were then registered with digitally‐reconstructed‐radiographs (DRR) obtained from treatment planning. Shifts between the kV images and reference DRR's were calculated: anterior‐posterior (A‐P), left‐right (L‐R) and superior‐inferior (S‐I). If the shifts were > 2 mm in any direction, the patient was reset within the immobilization mask until satisfying accuracy was achieved. Results: Shifts as large as 5 mm in the A‐P, L‐R and S‐I directions between kV projections and DRR's were found. These shifts represent offsets between the treatment and simulation setup using immobilization mask. The mean offsets of 0.1 mm, 0.7 mm, and −1.6 mm represent systematic errors in the A‐P, L‐R and S‐I directions, respectively, using the BrainLAB localizer. The mean of the radial shifts is about 1.7 mm. The standard deviations of shifts 2.2 mm, 2.0 mm, and 2.6 mm represent patient setup accuracy with the BrainLAB mask in A‐P, L‐R and S‐I directions. Conclusions: The BrainLAB mask provides less invasive immobilization system than the head ring. Relying on this system for patient setup might be associated with errors as large as 5 mm. Image‐guidance with the kV on‐board imager provides shifts that are required to setup patient prior to or during patient treatment. The patient may relax or move during treatment, and thus uncontrolled and undetected setup errors may be produced with patients that are not well immobilized. The combination of stereotactic immobilization and image guidance will achieve more controlled and accurate patient setup within 2 mm.