Engineering topics

Valdes, G

Publications and source records attributed to Valdes, G.

6 records

DOE OSTI2016

MO-FG-202-09: Virtual IMRT QA Using Machine Learning: A Multi-Institutional Validation

Purpose: To validate a machine learning approach to Virtual IMRT QA for accurately predicting gamma passing rates using different QA devices at different institutions. Methods: A Virtual IMRT QA was constructed using a machine learning algorithm based on 416 IMRT plans, in which QA measurements were performed using diode-array detectors and a 3%local/3mm with 10% threshold. An independent set of 139 IMRT measurements from a different institution, with QA data based on portal dosimetry using the same gamma index and 10% threshold, was used to further test the algorithm. Plans were characterized by 90 different complexity metrics. A weighted poison regression with Lasso regularization was trained to predict passing rates using the complexity metrics as input. Results: In addition to predicting passing rates with 3% accuracy for all composite plans using diode-array detectors, passing rates for portal dosimetry on per-beam basis were predicted with an error <3.5% for 120 IMRT measurements. The remaining measurements (19) had large areas of low CU, where portal dosimetry has larger disagreement with the calculated dose and, as such, large errors were expected. These beams need to be further modeled to correct the under-response in low dose regions. Important features selected by Lasso to predict gamma passing rates were: complete irradiated area outline (CIAO) area, jaw position, fraction of MLC leafs with gaps smaller than 20 mm or 5mm, fraction of area receiving less than 50% of the total CU, fraction of the area receiving dose from penumbra, weighted Average Irregularity Factor, duty cycle among others. Conclusion: We have demonstrated that the Virtual IMRT QA can predict passing rates using different QA devices and across multiple institutions. Prediction of QA passing rates could have profound implications on the current IMRT process.

60 APPLIED LIFE SCIENCES↗

DOE OSTI2016

TU-H-CAMPUS-JeP3-04: Factors Predicting a Need for Treatment Replanning with Proton Radiotherapy for Lung Cancer

Purpose: Proton dose distribution is sensitive to tumor regression and tissue and normal anatomy changes. Replanning is sometimes necessary during treatment to ensure continue tumor coverage or avoid overtreatment of organs at risk (OARs). We investigated action thresholds for replanning and identified both dosimetric and non-dosimetric metrics that would predict a need for replan. Methods: All consecutive lung cancer patients (n = 188) who received definitive proton radiotherapy and had more than two evaluation CT scans at the Roberts Proton Therapy Center (Philadelphia, USA) from 2011 to 2015 were included in this study. The cohort included a variety of tumor sizes, locations, histology, beam angles, as well as radiation-induced tumor and lung change. Dosimetric changes during therapy were characterized by changes in the dose volume distribution of PTV, ITV, and OARs (heart, cord, esophagus, brachial plexus and lungs). Tumor and lung change were characterized by changes in sizes, and in the distribution of Hounsfield numbers and water equivalent thickness (WET) along the beam path. We applied machine learning tools to identify both dosimetric and non-dosimetric metrics that predicted a replan. Results: Preliminary data showed that clinical indicators (n = 54) were highly correlated; thus, a simple indicator may be derived to guide the action threshold for replanning. Additionally, tumor regression alone could not predict dosimetric changes in OARs; it required further information about beam angles and tumor locations. Conclusion: Both dosimetric and non-dosimetric factors are predictive of the need for replanning during proton treatment.

60 APPLIED LIFE SCIENCES↗

DOE OSTI2015

SU-C-BRD-03: Closing the Loop On Virtual IMRT QA

Purpose: To develop an algorithm that predicts a priori IMRT QA passing rates. Methods: 416 IMRT plans from all treatment sites were planned in Eclipse version 11 and delivered using a dynamic sliding window technique on Clinac iX or TrueBeam linacs (Varian Medical Systems, Palo Alto, CA). The 3%/3mm and 2%/2mm local distance to agreement (DTA) were recorded during clinical operations using a commercial 2D diode array (MapCHECK 2, Sun Nuclear, Melbourne, FL). Each plan was characterized by 37 metrics that describe different failure modes between the calculated and measured dose. Machine-learning algorithms (MLAs) were trained to learn the relation between the plan characteristics and each passing rate. Minimization of the cross validated error, together with maximum a posteriori estimation (MAP), were used to choose the model parameters. Results: 3%/3mm local DTA can be predicted with an error smaller than 3% for 98% of the plans. For the remaining 2% of plans, the residual error was within 5%. For 2%/2mm local DTA passing rates, 96% percent of the plans were successfully predicted with an error smaller than 5%. All high-risk plans that failed the 2%/2mm local criteria were correctly identified by the algorithm. The most important metric to describe the passing rates was determined to be the MU per Gray (modulation factor). Conclusions: Logs files and independent dose calculations have been suggested as possible substitutes for measurement based IMRT QA. However, none of these methods answer the fundamental question of whether a plan can be delivered with a clinically acceptable error given the limitations of the linacs and the treatment planning system. Predicting the IMRT QA passing rates a priori closes that loop. For additional robustness, virtual IMRT QA can be combined with Linac QA and log file analysis to confirm appropriate delivery.

60 APPLIED LIFE SCIENCES↗

DOE OSTI2015

SU-E-J-45: The Correlation Between CBCT Flat Panel Misalignment and 3D Image Guidance Accuracy

Purpose To simulate the impact of CBCT flat panel misalignment on the image quality, the calculated correction vectors in 3D image guided proton therapy and to determine if these calibration errors can be caught in our QA process. Methods The X-ray source and detector geometrical calibration (flexmap) file of the CBCT system in the AdaPTinsight software (IBA proton therapy) was edited to induce known changes in the rotational and translational calibrations of the imaging panel. Translations of up to ±10 mm in the x, y and z directions (see supplemental) and rotational errors of up to ±3° were induced. The calibration files were then used to reconstruct the CBCT image of a pancreatic patient and CatPhan phantom. Correction vectors were calculated for the patient using the software’s auto match system and compared to baseline values. The CatPhan CBCT images were used for quantitative evaluation of image quality for each type of induced error. Results Translations of 1 to 3 mm in the x and y calibration resulted in corresponding correction vector errors of equal magnitude. Similar 10mm shifts were seen in the y-direction; however, in the x-direction, the image quality was too degraded for a match. These translational errors can be identified through differences in isocenter from orthogonal kV images taken during routine QA. Errors in the z-direction had no effect on the correction vector and image quality.Rotations of the imaging panel calibration resulted in corresponding correction vector rotations of the patient images. These rotations also resulted in degraded image quality which can be identified through quantitative image quality metrics. Conclusion Misalignment of CBCT geometry can lead to incorrect translational and rotational patient correction vectors. These errors can be identified through QA of the imaging isocenter as compared to orthogonal images combined with monitoring of CBCT image quality.

60 APPLIED LIFE SCIENCES↗

DOE OSTI2014

TU-F-17A-08: The Relative Accuracy of 4D Dose Accumulation for Lung Radiotherapy Using Rigid Dose Projection Versus Dose Recalculation On Every Breathing Phase

Purpose: To investigate the accuracy of 4D dose accumulation using projection of dose calculated on the end-exhalation, mid-ventilation, or average intensity breathing phase CT scan, versus dose accumulation performed using full Monte Carlo dose recalculation on every breathing phase. Methods: Radiotherapy plans were analyzed for 10 patients with stage I-II lung cancer planned using 4D-CT. SBRT plans were optimized using the dose calculated by a commercially-available Monte Carlo algorithm on the end-exhalation 4D-CT phase. 4D dose accumulations using deformable registration were performed with a commercially available tool that projected the planned dose onto every breathing phase without recalculation, as well as with a Monte Carlo recalculation of the dose on all breathing phases. The 3D planned dose (3D-EX), the 3D dose calculated on the average intensity image (3D-AVE), and the 4D accumulations of the dose calculated on the end-exhalation phase CT (4D-PR-EX), the mid-ventilation phase CT (4D-PR-MID), and the average intensity image (4D-PR-AVE), respectively, were compared against the accumulation of the Monte Carlo dose recalculated on every phase. Plan evaluation metrics relating to target volumes and critical structures relevant for lung SBRT were analyzed. Results: Plan evaluation metrics tabulated using 4D-PR-EX, 4D-PR-MID, and 4D-PR-AVE differed from those tabulated using Monte Carlo recalculation on every phase by an average of 0.14±0.70 Gy, - 0.11±0.51 Gy, and 0.00±0.62 Gy, respectively. Deviations of between 8 and 13 Gy were observed between the 4D-MC calculations and both 3D methods for the proximal bronchial trees of 3 patients. Conclusions: 4D dose accumulation using projection without re-calculation may be sufficiently accurate compared to 4D dose accumulated from Monte Carlo recalculation on every phase, depending on institutional protocols. Use of 4D dose accumulation should be considered when evaluating normal tissue complication probabilities as well as in clinical situations where target volumes are directly inferior to mobile critical structures.

60 APPLIED LIFE SCIENCES↗

DOE OSTI2014

SU-E-T-48: Automated Quality Assurance for XML Controlled Linacs

Purpose: To automate routine imaging QA procedures so that complying with TG 142 and TG 179 can be efficient and reliable. Methods: Two QA tests for a True Beam Linac were automatized. A Winston Lutz test as described by Lutz et al{sup 1} using the Winston Lutz test kit from BrainLab, Germany and a CBCT Image Quality test as described in TG 179 using the EMMA phantom, Siemens Medical Physics, Germany were performed in our True Beam. For each QA procedure tested, a 3 step paradigm was used. First, the data was automatically acquired using True Beam Developer Mode and XML scripting. Second, the data acquired in the first step was automatically processed using in-home grown Matlab GUIs. Third, Machine Learning algorithms were used to automatically classify the processed data and reports generated. Results: The Winston Luzt test could be performed by an experienced medical physicist in 29.0 ± 8.0 min. The same test, if automated using our paradigm, could be performed in 3.0 ± 0.1 min. In the same lieu, time could be substantially saved for image quality tests. In this case, the amount of time saved will depend on the phantoms used and the initial localization method. Additionally, machine learning algorithms could automatically identify the roots of the problems if any and possibly help reduce machine down time. Conclusion: Modern linear accelerators are equipped with advanced 2D and 3D imaging that are used for patient alignment substantially improving IGRT protocols. However, this extra complexity exponentially increases the number of QA tests needed. Using the new paradigm described above, not only bare minimum but best practice QA programs could be implemented with the same manpower. This work is supported by Varian, Palo Alto, CA.

60 APPLIED LIFE SCIENCES↗