Modeling the Agility MLC in the Monaco treatment planning system.

We investigate the relationship between the various parameters in the Monaco MLC model and dose calculation accuracy for an Elekta Agility MLC. The vendor-provided MLC modeling procedure - completed first with external vendor participation and then exclusively in-house - was used in combination with our own procedures to investigate several sets of MLC modeling parameters to determine their effect on dose distributions and point-dose measurements. Simple plans provided in the vendor procedure were used to elucidate specific mechanical characteristics of the MLC, while ten complex treatment plans - five IMRT and five VMAT - created using TG-119-based structure sets were used to test clinical dosimetric effects of particular parameter choices. EDR2 film was used for the vendor fields to give high spatial resolution, while a combination of MapCHECK and ion chambers were used for the in-house TG-119-based proced-ures. The vendor-determined parameter set provided a reasonable starting point for the MLC model and largely delivered acceptable gamma pass rates for clinical plans - including a passing external evaluation using the IROC H&N phantom. However, the vendor model did not provide point-dose accuracy consistent with that seen in other treatment systems at our center. Through further internal testing it was found that there existed many sets of MLC parameters, often at opposite ends of their allowable ranges, that provided similar dosimetric characteristics and good agreement with planar and point-dose measurements. In particular, the leaf offset and tip leakage parameters compensated for one another if adjusted in opposite directions, which provided a level curve of acceptable parameter sets across all plans. Interestingly, gamma pass rates of the plans were less dependent upon parameter choices than point-dose measurements, suggesting that MLC modeling using only gamma evaluation may be generally an insufficient approach. It was also found that exploring all parameters of the very robust MLC model to find the best match to the vendor-provided QA fields can reduce the pass rates of the TG-119-based clinical distributions as compared to simpler models. A wide variety of parameter sets produced MLC models capable of meeting RPC passing criteria for their H&N IMRT phantom. The most accurate models were achievable using a combination of vendor-provided and in-house procedures. The potential existed for an over-modeling of the Agility MLC in an effort to obtain the fine structure of certain quality assurance fields, which led to a reduction in agreement between calculation and measurement of more typical clinical dose distributions.

Voxel-based statistical analysis of uncertainties associated with deformable image registration.

Deformable image registration (DIR) algorithms have inherent uncertainties in their displacement vector fields (DVFs).The purpose of this study is to develop an optimal metric to estimate DIR uncertainties. Six computational phantoms have been developed from the CT images of lung cancer patients using a finite element method (FEM). The FEM generated DVFs were used as a standard for registrations performed on each of these phantoms. A mechanics-based metric, unbalanced energy (UE), was developed to evaluate these registration DVFs. The potential correlation between UE and DIR errors was explored using multivariate analysis, and the results were validated by landmark approach and compared with two other error metrics: DVF inverse consistency (IC) and image intensity difference (ID). Landmark-based validation was performed using the POPI-model. The results show that the Pearson correlation coefficient between UE and DIR error is rUE-error = 0.50. This is higher than rIC-error = 0.29 for IC and DIR error and rID-error = 0.37 for ID and DIR error. The Pearson correlation coefficient between UE and the product of the DIR displacements and errors is rUE-error × DVF = 0.62 for the six patients and rUE-error × DVF = 0.73 for the POPI-model data. It has been demonstrated that UE has a strong correlation with DIR errors, and the UE metric outperforms the IC and ID metrics in estimating DIR uncertainties. The quantified UE metric can be a useful tool for adaptive treatment strategies, including probability-based adaptive treatment planning.