No abstract

Three intensity-modulated tangential beam radiotherapy plan types for breast cancer treatment were evaluated based on PTV homogeneity index (HI) and equivalent uniform dose (EUD), heart V30 and EUD, whole lung V20 and EUD, and typical planning time compared to conventional 2D plans. 20 early-stage breast cancer patients were CT-scanned in the supine position, and tangential field extent, gantry and collimator angles were chosen. Four treatment plans were created for each patient: conventional, dynamically wedged plan based on the dose distribution on the central axial slice; forward planned IMRT; surface compensated plan created using an Eclipse tool and hybrid IMRT plan combining open and inverse-optimized fields. All three IMRT planning techniques represent significant improvement in PTV HI and EUD compared to conventional plans. Among the IMRT plans, the hybrid IMRT plan produced the best HI. IMRT lowered heart V30 and lung V20, but no significant differences in heart or lung EUD were detected between IMRT techniques. The IMRT technique with the shortest planning time was the compensated plan, followed by the hybrid IMRT. IMRT planning provides dosimetric benefits in breast cancer patients. The selection of the most appropriate IMRT technique must include careful consideration of the resources available.

The performance of a convolution/superposition based treatment planning system depends on the ability of the dose calculation algorithm to accurately account for physical interactions taking place in the tissue, key components of the linac head and on the accuracy of the photon beam model. Generally the user has little or no control over the performance of the dose calculation algorithm but is responsible for the accuracy of the beam model within the constraints imposed by the system. This study explores the dosimetric impact of limitations in photon beam modeling accuracy on complex 3D clinical treatment plans. A total of 70 photon beam models was created in the Pinnacle treatment planning system. Two of the models served as references for 6 MV and 15 MV beams, while the rest were created by perturbing the reference models in order to produce specific deviations in specific regions of the calculated dose profiles (central axis and transverse). The beam models were then used to generate 3D plans on seven CT data sets each for four different treatment sites (breast and conformal prostate, lung and brain). The equivalent uniform doses (EUD) of the targets and the principal organs at risk (OARs) of all plans ( approximately 1000) were calculated and compared to the EUDs delivered by the reference beam models. In general, accurate dosimetry of the target is most greatly compromised by poor modeling of the central axis depth dose and the horns, while the EUDs of the OARs exhibited the greatest sensitivity to beam width accuracy. Based on the results of this analysis we suggest a set of tolerances to be met during commissioning of the beam models in a treatment planning system that are consistent in terms of clinical outcomes as predicted by the EUD.

Purpose Increased use of Linac‐based stereotactic radiosurgery (SRS), which requires highly noncoplanar gantry trajectories, necessitates the development of efficient and accurate methods of collision detection during the treatment planning process. This work outlines the development and clinical implementation of a patient‐specific computed tomography (CT) contour‐based solution that utilizes Eclipse Scripting to ensure maximum integration with clinical workflow. Methods The collision detection application uses triangle mesh structures of the gantry and couch, in addition to the body contour of the patient taken during CT simulation, to virtually simulate patient treatments. Collision detection is performed using Binary Tree Hierarchy detection methods. Algorithm accuracy was first validated for simple cuboidal geometry using a calibration phantom and then extended to an anthropomorphic phantom simulation by comparing the measured minimum distance between structures to the predicted minimum distance for all allowable orientations. The collision space was tested at couch angles every 15° from 90 to 270 with the gantry incremented by 5° through the maximum trajectory. Receiver operating characteristic curve analysis was used to assess algorithm sensitivity and accuracy for predicting collision events. Following extensive validation, the application was implemented clinically for all SRS patients. Results The application was able to predict minimum distances between structures to within 3 cm. A safety margin of 1.5 cm was sufficient to achieve 100% sensitivity for all test cases. Accuracy obtained was 94.2% with the 5 cm clinical safety margin with 100% true positive collision detection. A total of 88 noncoplanar SRS patients have been currently tested using the application with one collision detected and no undetected collisions occurring. The average time for collision testing per patient was 2 min 58 s. Conclusions A collision detection application utilizing patient CT contours was developed and successfully clinically implemented. This application allows collisions to be detected early during the planning process, avoiding patient delays and unnecessary resource utilization if detected during delivery.