Comparisons of treatment optimization directly incorporating random patient setup uncertainty with a margin-based approach.

The purpose of this study is to incorporate the dosimetric effect of random patient positioning uncertainties directly into a commercial treatment planning system's IMRT plan optimization algorithm through probabilistic treatment planning (PTP) and compare coverage of this method with margin-based planning. In this work, PTP eliminates explicit margins and optimizes directly on the estimated integral treatment dose to determine optimal patient dose in the presence of setup uncertainties. Twenty-eight prostate patient plans adhering to the RTOG-0126 criteria are optimized using both margin-based and PTP methods. Only random errors are considered. For margin-based plans, the planning target volume is created by expanding the clinical target volume (CTV) by 2.1 mm to accommodate the simulated 3 mm random setup uncertainty. Random setup uncertainties are incorporated into IMRT dose evaluation by convolving each beam's incident fluence with a sigma = 3 mm Gaussian prior to dose calculation. PTP optimization uses the convolved fluence to estimate dose to ensure CTV coverage during plan optimization. PTP-based plans are compared to margin-based plans with equal CTV coverage in the presence of setup errors based on dose-volume metrics. The sensitivity of the optimized plans to patient-specific setup uncertainty variations is assessed by evaluating dose metrics for dose distributions corresponding to halving and doubling of the random setup uncertainty used in the optimization. Margin-based and PTP-based plans show similar target coverage. A physician review shows that PTP is preferred for 21 patients, margin-based plans are preferred in 2 patients, no preference is expressed for 1 patient, and both autogenerated plans are rejected for 4 patients. For the PTP-based plans, the average CTV receiving the prescription dose decreases by 0.5%, while the mean dose to the CTV increases by 0.7%. The CTV tumor control probability (TCP) is the same for both methods with the exception of one case in which PTP gave a slightly higher TCP. For critical structures that do not meet the optimization criteria, PTP shows a decrease in the volume receiving the maximum specified dose. PTP reduces local normal tissue volumes receiving the maximum dose on average by 48%. PTP results in lower mean dose to all critical structures for all plans. PTP results in a 2.5% increase in the probability of uncomplicated control (P+), along with a 1.9% reduction in rectum normal tissue complication probability (NTCP), and a 0.7% reduction in bladder NTCP. PTP-based plans show improved conformality as compared with margin-based plans with an average PTP-based dosimetric margin at 7100 cGy of 0.65 cm compared with the margin-based 0.90 cm and a PTP-based dosimetric margin at 3960 cGy of 1.60 cm compared with the margin-based 1.90 cm. PTP-based plans show similar sensitivity to variations of the uncertainty during treatment from the uncertainty used in planning as compared to margin-based plans. For equal target coverage, when compared to margin-based plans, PTP results in equal or lower doses to normal structures. PTP results in more conformal plans than margin-based plans and shows similar sensitivity to variations in uncertainty.

Monte Carlo-based adaptive EPID dose kernel accounting for different field size responses of imagers.

The aim of this study is to present an efficient method to generate imager-specific Monte Carlo (MC)-based dose kernels for amorphous silicon-based electronic portal image device dose prediction and determine the effective backscattering thicknesses for such imagers. EPID field size-dependent responses were measured for five matched Varian accelerators from three institutions with 6 MV beams at the source to detector distance (SDD) of 105 cm. For two imagers, measurements were made with and without the imager mounted on the robotic supporting arm. Monoenergetic energy deposition kernels with 0-2.5 cm of water backscattering thicknesses were simultaneously computed by MC to a high precision. For each imager, the backscattering thickness required to match measured field size responses was determined. The monoenergetic kernel method was validated by comparing measured and predicted field size responses at 150 cm SDD, 10 x 10 cm2 multileaf collimator (MLC) sliding window fields created with 5, 10, 20, and 50 mm gaps, and a head-and-neck (H&N) intensity modulated radiation therapy (IMRT) patient field. Field size responses for the five different imagers deviated by up to 1.3%. When imagers were removed from the robotic arms, response deviations were reduced to 0.2%. All imager field size responses were captured by using between 1.0 and 1.6 cm backscatter. The predicted field size responses by the imager-specific kernels matched measurements for all involved imagers with the maximal deviation of 0.34%. The maximal deviation between the predicted and measured field size responses at 150 cm SDD is 0.39%. The maximal deviation between the predicted and measured MLC sliding window fields is 0.39%. For the patient field, gamma analysis yielded that 99.0% of the pixels have gamma < 1 by the 2%, 2 mm criteria with a 3% dose threshold. Tunable imager-specific kernels can be generated rapidly and accurately in a single MC simulation. The resultant kernels are imager position independent and are able to predict fields with varied incident energy spectra and a H&N IMRT patient field. The proposed adaptive EPID dose kernel method provides the necessary infrastructure to build reliable and accurate portal dosimetry systems.