A model-based 3D patient-specific pre-treatment QA method for VMAT using the EPID.

This study reports the development and validation of a model-based, 3D patient dose reconstruction method for pre-treatment quality assurance using EPID images. The method is also investigated for sensitivity to potential MLC delivery errors. Each cine-mode EPID image acquired during plan delivery was processed using a previously developed back-projection dose reconstruction model providing a 3D dose estimate on the CT simulation data. Validation was carried out using 24 SBRT-VMAT patient plans by comparing: (1) ion chamber point dose measurements in a solid water phantom, (2) the treatment planning system (TPS) predicted 3D dose to the EPID reconstructed 3D dose in a solid water phantom, and (3) the TPS predicted 3D dose to the EPID and our forward predicted reconstructed 3D dose in the patient (CT data). AAA and AcurosXB were used for TPS predictions. Dose distributions were compared using 3%/3 mm (95% tolerance) and 2%/2 mm (90% tolerance) γ-tests in the planning target volume (PTV) and 20% dose volumes. The average percentage point dose differences between the ion chamber and the EPID, AcurosXB, and AAA were 0.73 ± 1.25%, 0.38 ± 0.96% and 1.06 ± 1.34% respectively. For the patient (CT) dose comparisons, seven (3%/3 mm) and nine (2%/2 mm) plans failed the EPID versus AAA. All plans passed the EPID versus Acuros XB and the EPID versus forward model γ-comparisons. Four types of MLC sensitive errors (opening, shifting, stuck, and retracting), of varying magnitude (0.2, 0.5, 1.0, 2.0 mm), were introduced into six different SBRT-VMAT plans. γ-comparisons of the erroneous EPID dose and original predicted dose were carried out using the same criteria as above. For all plans, the sensitivity testing using a 3%/3 mm γ-test in the PTV successfully determined MLC errors on the order of 1.0 mm, except for the single leaf retraction-type error. A 2%/2 mm criteria produced similar results with two more additional detected errors.

An in vivo dose verification method for SBRT-VMAT delivery using the EPID.

PURPOSE: Radiation treatments have become increasingly more complex with the development of volumetric modulated arc therapy (VMAT) and the use of stereotactic body radiation therapy (SBRT). SBRT involves the delivery of substantially larger doses over fewer fractions than conventional therapy. SBRT-VMAT treatments will strongly benefit from in vivo patient dose verification, as any errors in delivery can be more detrimental to the radiobiology of the patient as compared to conventional therapy. Electronic portal imaging devices (EPIDs) are available on most commercial linear accelerators (Linacs) and their documented use for dosimetry makes them valuable tools for patient dose verification. In this work, the authors customize and validate a physics-based model which utilizes on-treatment EPID images to reconstruct the 3D dose delivered to the patient during SBRT-VMAT delivery. METHODS: The SBRT Linac head, including jaws, multileaf collimators, and flattening filter, were modeled using Monte Carlo methods and verified with measured data. The simulation provides energy spectrum data that are used by their "forward" model to then accurately predict fluence generated by a SBRT beam at a plane above the patient. This fluence is then transported through the patient and then the dose to the phosphor layer in the EPID is calculated. Their "inverse" model back-projects the EPID measured focal fluence to a plane upstream of the patient and recombines it with the extra-focal fluence predicted by the forward model. This estimate of total delivered fluence is then forward projected onto the patient's density matrix and a collapsed cone convolution algorithm calculates the dose delivered to the patient. The model was tested by reconstructing the dose for two prostate, three lung, and two spine SBRT-VMAT treatment fractions delivered to an anthropomorphic phantom. It was further validated against actual patient data for a lung and spine SBRT-VMAT plan. The results were verified with the treatment planning system (TPS) (ECLIPSE AAA) dose calculation. RESULTS: The SBRT-VMAT reconstruction model performed very well when compared to the TPS. A stringent 2%/2 mm χ-comparison calculation gave pass rates better than 91% for the prostate plans, 88% for the lung plans, and 86% for the spine plans for voxels containing 80% or more of the prescribed dose. Patient data were 86% for the lung and 95% for the spine. A 3%/3 mm χ-comparison was also performed and gave pass rates better than 93% for all plan types. CONCLUSIONS: The authors have customized and validated a robust, physics-based model that calculates the delivered dose to a patient for SBRT-VMAT delivery using on-treatment EPID images. The accuracy of the results indicates that this approach is suitable for clinical implementation. Future work will incorporate this model into both offline and real-time clinical adaptive radiotherapy.

Poster - Thur Eve - 25: In vivo dosimetric verification of intensity-modulated radiation therapy.

Dosimetric verification of patient treatment plans has become increasingly important due to the widespread use of complicated delivery techniques. IMRT and VMAT treatments are typically verified prior to start of the patient's course of treatment, using a point dose and/or a film measurement. Pre-treatment verification will not detect patient or machine-related errors; therefore, in vivo dosimetric verification is the only way to determine if the patient's treatment was delivered correctly. Portal images were acquired throughout the course of five prostate and six head-and-neck patient IMRT treatments. The corresponding predicted images were calculated using a previously developed portal dose image prediction algorithm, which combines a versatile fluence model with a patient scatter and EPID dose prediction model. The prostate patient image agreement was found to vary day-to-day due to rectal gas pockets and the effect of adjustable support rails on the patient couch. The head-and-neck patient images were observed to be more consistent daily, but an increased measured dose was evident at the periphery of the patient, likely due to patient weight loss. The majority of the fields agreed within 3% and 3 mm for greater than 90% of the pixels, as established by the χ-comparison. This work demonstrates the changes in patient anatomy that are detectable with the portal dose image prediction model. Prior to clinical implementation, the effect of the couch must be incorporated into the model, the image acquisition must be automatically scheduled and routine EPID QA must be undertaken to ensure the collection of high-quality EPID images.

Poster - Thur Eve - 51: Three-dimensional in-vivo EPID dosimetry of IMRT and VMAT treatments.

As radiation treatment delivery becomes more complex, including dynamic IMRT and VMAT, the argument for routine patient dose verification becomes more compelling. This work demonstrates a technique that utilizes our pre-existing portal dose image prediction algorithm to compute 3D patient dose from recorded on-treatment portal images. This approach can be applied on CT simulation data or daily cone-beam CT data sets. Here we demonstrate the robustness of our dose reconstruction technique with phantom and patient examples, with delivery schemes including IMRT and VMAT. For an example prostate treatment site, 3D dose distributions reconstructed in the patient model are computed for each fraction, and DVHs presented. Results indicate that the patient dose reconstruction algorithm compares well with treatment planning system computed doses for controlled test situations. For patient examples the 3D chi comparison values (similar to the gamma comparison) ranged from 94.5% to 100% agreement for voxels > 10% maximum dose for all treatments and phantom cases. We show an example where the DVH for fraction nine of a prostate treatment fails acceptability criteria, due to a previously unnoticed positioning error. Future work involves building our patient dose reconstruction into a QA package, subsequently integrating it into a clinical workflow. We are also investigating the use of this tool as a backbone for an in-house adaptive radiotherapy implementation. This work is supported by Varian Medical Systems.