Armia George

Rapid learning in practice: A lung cancer survival decision support system in routine patient care data

Background and purpose A rapid learning approach has been proposed to extract and apply knowledge from routine care data rather than solely relying on clinical trial evidence. To validate this in practice we deployed a previously developed decision support system (DSS) in a typical, busy clinic for non-small cell lung cancer (NSCLC) patients. Material and methods Gender, age, performance status, lung function, lymph node status, tumor volume and survival were extracted without review from clinical data sources for lung cancer patients. With these data the DSS was tested to predict overall survival. Results 3919 lung cancer patients were identified with 159 eligible for inclusion, due to ineligible histology or stage, non-radical dose, missing tumor volume or survival. The DSS successfully identified a good prognosis group and a medium/poor prognosis group (2 year OS 69% vs. 27/30%, p < 0.001). Stage was less discriminatory (2 year OS 47% for stage I–II vs. 36% for stage IIIA–IIIB, p = 0.12) with most good prognosis patients having higher stage disease. The DSS predicted a large absolute overall survival benefit (~40%) for a radical dose compared to a non-radical dose in patients with a good prognosis, while no survival benefit of radical radiotherapy was predicted for patients with a poor prognosis. Conclusions A rapid learning environment is possible with the quality of clinical data sufficient to validate a DSS. It uses patient and tumor features to identify prognostic groups in whom therapy can be individualized based on predicted outcomes. Especially the survival benefit of a radical versus non-radical dose predicted by the DSS for various prognostic groups has clinical relevance, but needs to be prospectively validated.

Is a quasi-3D dosimeter better than a 2D dosimeter for Tomotherapy delivery quality assurance?

Delivery quality assurance (DQA) has been performed for each Tomotherapy patient either using ArcCHECK or MatriXX Evolution in our clinic since 2012. ArcCHECK is a quasi-3D dosimeter whereas MatriXX is a 2D detector. A review of DQA results was performed for all patients in the last three years, a total of 221 DQA plans. These DQA plans came from 215 patients with a variety of treatment sites including head-neck, pelvis, and chest wall. The acceptable Gamma pass rate in our clinic is over 95% using 3mm and 3% of maximum planned dose with 10% dose threshold. The mean value and standard deviation of Gamma pass rates were 98.2% ± 1.98(1SD) for MatriXX and 98.5%±1.88 (1SD) for ArcCHECK. A paired t-test was also performed for the groups of patients whose DQA was performed with both the ArcCHECK and MatriXX. No statistical dependence was found in terms of the Gamma pass rate for ArcCHECK and MatriXX. The considered 3D and 2D dosimeters have achieved similar results in performing routine patient-specific DQA for patients treated on a TomoTherapy unit.

Evaluation of 3D Gamma index calculation implemented in two commercial dosimetry systems

3D Gamma index is one of the metrics which have been widely used for clinical routine patient specific quality assurance for IMRT, Tomotherapy and VMAT. The algorithms for calculating the 3D Gamma index using global and local methods implemented in two software tools: PTW- VeriSoft® as a part of OCTIVIUS 4D dosimeter systems and 3DVHTM from Sun Nuclear were assessed. The Gamma index calculated by the two systems was compared with manual calculated for one data set. The Gamma pass rate calculated by the two systems was compared using 3%/3mm, 2%/2mm, 3%/2mm and 2%/3mm for two additional data sets. The Gamma indexes calculated by the two systems were accurate, but Gamma pass rates calculated by the two software tools for same data set with the same dose threshold were different due to the different interpolation of raw dose data by the two systems and different implementation of Gamma index calculation and other modules in the two software tools. The mean difference was -1.3%±3.38 (1SD) with a maximum difference of 11.7%.

SU‐E‐T‐407: Evaluation of Four Commercial Dosimetry Systems for Routine Patient‐Specific Tomotherapy Delivery Quality Assurance

PURPOSE The purpose of this project was to evaluate the performance of four commercially available dosimetry systems for Tomotherapy delivery quality assurance (DQA). METHODS Eight clinical patient plans were chosen to represent a range of treatment sites and typical clinical plans. Four DQA plans for each patient plan were created using the TomoTherapy DQA Station (Hi-Art version 4.2.1) on CT images of the ScandiDose Delta4, IBA MatriXX Evolution, PTW Octavius 4D and Sun Nuclear ArcCHECK phantoms. Each detector was calibrated following the manufacture-provided procedure. No angular response correction was applied. All DQA plans for each detector were delivered on the Tomotherapy Hi-Art unit in a single measurement session but on different days. The measured results were loaded into the vendor supplied software for each QA system for comparison with the TPS-calculated dose. The Gamma index was calculated using 3%/3mm, 2%/2mm with 10% dose threshold of maximum TPS calculated dose. RESULTS Four detector systems showed comparable gamma pass rates for 3%/3m, which is recommended by AAPM TG119 and commonly used within the radiotherapy community. The averaged pass rates ± standard deviation for all DQA plans were (98.35±1.97)% for ArcCHECK, (99.9%±0.87)% for Matrix, (98.5%±5.09)% for Octavius 4D, (98.7%±1.27)% for Delata4. The rank of the gamma pass rate for individual plans was consistent between detectors. Using 2%/2mm Gamma criteria for analysis, the Gamma pass rate decreased on average by 9%, 8%, 6.6% and 5% respectively. Profile and Gamma failure map analysis using the software tools from each dosimetry system indicated that decreased passing rate is mainly due to the threading effect of Tomo plan. CONCLUSION Despite the variation in detector type and resolution, phantom geometry and software implementation, the four systems demonstrated similar dosimetric performance, with the rank of the gamma pass rate consistent for the plans considered.

Technical Note: Penumbral width trimming in solid lung dose profiles for 0.9 and 1.5 T MRI‐Linac prototypes

PURPOSE Longitudinal magnetic fields narrow beam penumbra and tighten lateral spread of secondary electrons in air cavities, including lung tissue. Gafchromic® EBT3 film was used to investigate differences between penumbra in solid water and solid lung, without a magnetic field (0 T) and with two field strengths (0.9 and 1.5 T). METHODS The first prototype of the Australian MRI-linac consisted of a 1.5 T Siemens Sonata MRI and Varian industrial linatron (nominal 4 MV). The second prototype replaced the Sonata with a 1.0 T Agilent split-bore magnet. Measurements were completed at 0.9 T to maintain the same source-to-surface distance between set-ups. Gammex-rmi® solid water with 50 mm of CIRS solid lung inserted as a lung cavity was positioned inside each magnet. This was compared to the same set-up with solid water only, where film measurements were completed at solid water equivalent depths corresponding to entrance interface/mid/exit interface positions of solid lung from the first set-up. Multileaf collimator (MLC)-defined field sizes were set to 3 × 3 cm2 and 10 × 10 cm2 . The 80%-20% penumbral width was determined. RESULTS Under 1.5 T conditions, penumbra narrowing occurred up to 4.4 ± 0.1 mm compared to 0 T. As expected, the effect was less for 0.9 T, which resulted in a maximum narrowing of 2.5 ± 0.1 mm. Exit profile penumbra were more affected than entrance penumbra by up to 2.6 ± 0.2 mm. The 1.5 T field brought the solid water and lung penumbral widths more into alignment by a maximum difference of 0.4 ± 0.1 mm. CONCLUSIONS The trimming of penumbral widths due to magnetic fields in solid water and lung was demonstrated and compared to 0 T. The 0.9 and 1.5 T field trimmed the penumbra by up to 2.5 ± 0.1 mm and 4.4 ± 0.1 mm respectively.

The Australian MRI-Linac Program: measuring profiles and PDD in a horizontal beam

The Australian MRI-Linac consists of a fixed horizontal photon beam combined with a MRI. Commissioning required PDD and profiles measured in a horizontal set-up using a combination of water tank measurements and gafchromic film. To validate the methodology, measurements were performed comparing PDD and profiles measured with the gantry angle set to 0 and 90° on a conventional linac. Results showed agreement to within 2.0% for PDD measured using both film and the water tank at gantry 90° relative to PDD acquired using gantry 0°. Profiles acquired using a water tank at both gantry 0 and 90° showed agreement in FWHM to within 1 mm. The agreement for both PDD and profiles measured at gantry 90° relative to gantry 0° curves indicates that the methodology described can be used to acquire the necessary beam data for horizontal beam lines and in particular, commissioning the Australian MRI-linac.

SU‐E‐T‐503: Development of a Software Tool for Verification of Delivered Tomotherapy Plans Using the Tomo Log File

PURPOSE The data receive server (DRS) in the Tomotherapy Unit records planned and actually delivered plan parameters for each treatment into a log file. The purpose of this study was to develop a software tool using the log file for verification of the patient plan delivered during treatment. METHODS A software tool, TomoPQA, was developed using the Python programming language. The software was implemented using object-oriented methodology and modular design. The program has three built-in modules: Read-in module for loading the log file, analysis module for analyzing the log file and reporting module for producing a PDF report. RESULTS The developed software tool can be used to monitor and check the following plan parameters during patient treatment: (1) planned and actual field width; (2) planned and actual treatment time; (3) planned and actual couch speed; (4) planned and actual gantry speed; (4) planned and actual setup position. The software shows the difference between these parameters as a graphic plot against each treatment fraction in a PDF report for easy review or these values can be exported to an excel file. The program can process a 100M byte log file and produce a report in the order of one minute and can run on Windows, Linux or Mac platforms as a standalone program. A server version of this program can also be implemented for full automation of the log file processing, generation of the PDF report and also has the potential for an automated email if given values are out of tolerance. CONCLUSION A QA software tool has been developed for in-vivo quality assurance of treatment delivered on a Tomotherapy Unit. This tool provides an independent verification of the difference between actual delivered and planned parameters during treatment.

Commissioning of SharePlan: The Liverpool Experience

SharePlan is a treatment planning system developed by Raysearch Laboratories AB to enable creation of a linear accelerator intensity modulated radiotherapy (IMRT) plan as a backup for a Tomotherapy plan. A 6MV Elekta Synergy Linear accelerator photon beam was modelled in SharePlan. The beam model was validated using Matrix Evolution, a 2D ion chamber array, for two head-neck and three prostate plans using 3%/3mm Gamma criteria. For 39 IMRT beams, the minimum and maximum Gamma pass rates are 95.4% and 98.7%. SharePlan is able to generate backup IMRT plans which are deliverable on a traditional linear accelerator and accurate in terms of clinical criteria. During use of SharePlan, however, an out-of-memory error frequently occurred and SharePlan was forced to be closed. This error occurred occasionally at any of these steps: loading the Tomotherapy plan into SharePlan, generating the IMRT plan, selecting the optimal plan, approving the plan and setting up a QA plan. The out-of-memory error was caused by memory leakage in one or more of the C/C++ functions implemented in SharePlan fluence engine, dose engine or optimizer, as acknowledged by the manufacturer. Because of the interruption caused by out-of-memory errors, SharePlan has not been implemented in our clinic although accuracy has been verified. A new software program is now being provided to our centre to replace SharePlan.