Poster - Thur Eve - 28: Optimization of a prostate cancer IGRT protocol.

Our image-guided radiation therapy (IGRT) protocol for post-prostatectomy patients involves acquiring a kV cone beam computed tomography (CBCT) dataset at each fraction and shifting the treatment couch to align the surgical clips. This IGRT strategy requires significant resources, and delivers non-negligible dose to normal tissues. The objective of this work is to evaluate this IGRT protocol against two alternative strategies in terms of the dose-volume statistics for target and organ at risk regions. Our method involves deforming the planning CT to the CBCT dataset acquired at each fraction, computing dose on the deformed dataset, and inversely transforming the dose back onto the original planning CT dataset. The treatments of six patients were evaluated assuming three IGRT scenarios: no IGRT, daily IGRT using the clinically employed couch shifts, and alternating day IGRT. The doses delivered to the clinical target volumes are within approximately 3.2, 1.3, and 2.1% of the plan for the non-IGRT, daily, and alternating day IGRT protocols, respectively. Doses to relevant portions of the organs at risk deviate from the plan by up to 10.5, 13.1 and 10.7% for non-IGRT, daily IGRT, and alternating day IGRT protocols, respectively. Some cases do not differ significantly between IGRT and non-IGRT protocols in terms of cumulative DVHs, highlighting the difficult task of correcting prostate bed deformations via the treatment couch translations. In general, the alternating day IGRT protocol was found to result in a clinically insignificant deviation in delivered dose while providing a significant reduction in resource use and patient imaging dose.

Poster - Thurs Eve-26: Influence of MLC leaf edge and tongue and groove effect on IMRT dose distributions.

An important consideration when using multileaf collimators (MLCs) for IMRT delivery is the correct account of leaf edge effects. To study these effects, two maps were constructed from a clinical beam's step and shoot delivery sequence: a time-weighted leaf edge position (LEP) map where the pixel intensity was proportional to the length of time that a leaf edge defined the edge of any segment within the field, and a (TG) map where pixel intensity was proportional to the length of time that adjacent segments matched along a leaf edge. We investigated the correlation between LEP or TG maps with dose error maps (obtained by subtracting calculated from either measured (CM) or from re-calculated (CC) data). Re-calculated data were obtained by modifying selected MLC photon modeling parameters from their commissioned values. We calculated the correlation coefficient between corresponding regions of CM and CC maps with TG and LEP maps. A NAT analysis of the CM maps indicated that the NAT index was minimized for tongue and groove width at the commissioned value of 0.1cm. A higher correlation coefficient was seen between CM and LEP maps (0.62±0.11) than between CM or CC and TG maps across all MLC modeling parameters used. The low correlation between both LEP or TG maps and CC maps suggests that the higher correlation observed between both LEP or TG maps and CM maps cannot be attributed to the choice of MLC modeling parameters alone. Further work is needed to pinpoint the cause of this correlation.

Poster - Thurs Eve-36: Use of multileaf collimator as a replacement of physical missing tissue compensator.

Missing tissue compensators are used to improve dose uniformity for some patients undergoing radiation therapy. Currently, our practice is to machine compensators out of lead alloy plate. Replacing this physical filter with a segmented multileaf collimator (MLC) delivery sequence is beneficial in terms of work flow and delivery efficiency. The purpose of this work is to compare the dose uniformity achieved by fields that are either (A) conventionally compensated, compensated by segmenting the physical compensator thickness map into either (B) step-and-shoot or (C) dynamic MLC delivery sequences using an in-house sequencer, (D) compensated using Pinnacle sequencer, or (E) compensated using IMRT optimization. A computer program was developed to construct both step-and-shoot and dynamic MLC sequence files from mechanical thickness maps of our current compensators. In addition, the Pinnacle sequencer and IMRT optimization were used to generate step-and-shoot MLC sequences. Planar doses were measured for each at the isocenter depth with an ion chamber array to compare the five methods. A comparison of the relative dose distribution shows that the compensation achieved by method (E) is in close agreement with that achieved using method (A), that is, dose uniformity within 4%. Method (D) resulted in the shortest delivery time and achieved dose uniformity to within 5%. Methods (B) and (C) need additional refinement to be of practical use. The results support the feasibility of replacing physical compensators with MLC delivery sequences. Compensation by MLC segments provides more flexibility and efficiency in design and delivery than by physical compensators while maintaining or improving the uniformity of dose to the plane of compensation.

Poster - Thurs Eve-34: Extended CT-range in RT planning of pelvic cancer treatment in presence of hip replacements.

Extended CT range in conventional CT scanners has a potential to allow for a more conformal treatment of patients with hip prosthesis. Its use may limit inaccuracies in electron density maps that are observed due to severe artifacts in CT data. In this study, we investigate the use of CT images with extended CT numbers in dose calculations and compare the results of calculations with standard CT data and measured doses. A phantom containing a hip prosthesis was scanned and treatment was planned with extended and standard CT numbers. The density override function was used to eliminate the effect of artifacts in the region outside of the metallic implant, while raw CT numbers were used to indicate density within the implant. Dose measurements were performed with two types of ion chambers at 6, 10 and 18MV energies. Our results indicate that data with extended CT range result in a better agreement between measured and calculated dose at the central position of the body phantom, as should be expected. However, CT artifacts within the implant region also result in discrepancies between the measured and calculated dose. The discrepancy is greater at lower cross-sectional thickness where bright, high density surface artifacts are high relative to the artificially low density inner region of the implant. Potential ways of resolving the discrepancies are outlined and a possibility of their application to clinical routine will be discussed.

Megavoltage imaging with low Z targets: implementation and characterization of an investigational system.

The poor quality of stereotactic radiotherapy portal images is a limiting factor in precise image registration. To alleviate this problem, a low atomic number (Z) target was implemented on our Siemens MXE linear accelerator. This investigational system was used to assess the performance of various target materials by filming an aluminum contrast object. Beryllium, carbon and conventional target materials were studied. The bremsstrahlung spectra of these materials were simulated using Monte Carlo techniques. These spectra were used to calculate the dependence of narrow beam contrast on phantom thickness for verification of the data measured from film. A Monte Carlo simulation of the beryllium spectrum in a wide beam geometry was used to evaluate the effect of phantom-to-film distance on contrast. Although the same degree of contrast improvement with distance was not realized in practice, the improvement in image quality rivaled that achieved using a scatter reduction grid. A comparison of conventional localization images of the head and neck of an anthropomorphic phantom with images produced with a beryllium or carbon target and a mammography film and screen system supports earlier suggestions that the technique is clinically useful.

Refinements of the finite-size pencil beam model of three-dimensional photon dose calculation.

Modern three-dimensional (3-D) photon dose calculation algorithms need to be fast and accurate if they are to be practical for treatment optimization. Refinements to a previously proposed finite-size pencil beam (FSPB) method are presented in order to fulfill these needs. Specifically, a fast Fourier transform (FFT) convolution technique is used to speed calculation of the FSPB; the fluence spectrum is modeled, and the effects of finite source size, a Gaussian x-ray source intensity profile and partial transmission through a multileaf collimator (MLC) leaf are approximated. The use of FFT techniques in the calculation of small diverging fields involves approximations that are investigated for a 6 MV beam and shown to introduce errors that vary with energy but do not exceed 0.7% on the central axis. Dose distributions calculated by FSPB superposition are in excellent agreement with those calculated by full field FFT convolution. Two key advances over the original implementation of the FSPB model are demonstrated: the fast calculation of the FSPB facilitates development, and the incorporation of realistic beam parameters enables accurate modeling of clinical beams.