Development of an Electronic Portal Imaging Device Dosimetry Method

Support arm backscatter and off-axis effects of an electronic portal imaging device (EPID) are challenging for radiotherapy quality assurance. Aiming at the issue, we proposed a simple yet effective method with correction matrices to rectify backscatter and off-axis responses for EPID images. First, we measured the square fields with ionization chamber array (ICA) and EPID simultaneously. Second, we calculated the dose-to-pixel value ratio and used it as the correction matrix of the corresponding field. Third, the correction value of the large field was replaced with that of the same point in the small field to generate a correction matrix suitable for different EPID images. Finally, we rectified the EPID image with the correction matrix, and then the processed EPID images were converted into the absolute dose. The calculated dose was compared with the measured dose via ICA. The gamma pass rates of 3%/3 mm and 2%/2 mm (5% threshold) were 99.6% ± 0.94% and 95.48% ± 1.03%, and the average gamma values were 0.28 ± 0.04 and 0.42 ± 0.05, respectively. Experimental results verified our method accurately corrected EPID images and converted pixel values into absolute dose values such that EPID was an efficient radiotherapy dosimetry tool.

Analysis of the Physical and Radiobiological Equivalence of the Calculated and Measured Dose Distributions for Prostate Stereotactic Radiotherapy

Purpose: Carrying out the analysis of the physical and radiobiological equivalence of dose distributions obtained during the planning of hypofractionated stereotactic radiation therapy of the prostate cancer and verification using a three-dimensional cylindrical dosimeter. Material and Methods: Based on the anatomical data of twelve patients diagnosed with prostate carcinoma, stage T2N0M0 with low risk, plans were developed for stereotactic radiation therapy with volumetric modulates arc therapy (VMAT). The dose per fraction was 7,25 Gy for 5 fractions (total dose 36,25 Gy) with a normal photon energy of 10 MV. The developed plans were verified using a three-dimensional cylindrical ArcCHECK phantom. During the verification process, the three-dimensional dose distribution in the phantom was measured, based on which the values of the three-dimensional gamma index and the dose–volume histogram within each contoured anatomical structures were calculated with 3DVH software. The gamma index value γ (3 %, 2 mm, GN) at a threshold equal to 20 % of the dose maximum of the plan and the percentage of coincidence of points at least 95 % was chosen as a criterion of physical convergence of the calculated and measured dose distribution according to the recommendations of AAPM TG-218. To analyze the radiobiological equivalence of the calculated and measured dose distribution, the local control probability (TCP) and normal tissue complication probability (NTCP) criteria were used based on the calculated and measured dose–volume histograms. Contours of the target (PTV) and the anterior wall of the rectum were used for the analysis. The approach based on the concept of equivalent uniform dose (EUD) by A. Niemierko was used to calculate the values of TCP/NTCP criteria. Results: The results of physical convergence of plans for all patients on the contour of the whole body were higher than 95 % for the criteria γ (3 %, 2 mm, GN). The convergence along the PTV contour is in the range (75.5–95.2)%. The TCP and NTCP values obtained from the measured dose-volume histograms were higher than the planned values for all patients. It was found that the accelerator delivered a slightly higher dose to the PTV and the anterior wall of the rectum than originally planned. Conclusion: The capabilities of modern dosimetric equipment allow us move to the verification of treatment plans based on the analysis of TCP / NTCP radiobiological equivalence, taking into account the individual characteristics of the patient and the capabilities of radiation therapy equipment.

DeepBeam: a machine learning framework for tuning the primary electron beam of the PRIMO Monte Carlo software

Background Any Monte Carlo simulation of dose delivery using medical accelerator-generated megavolt photon beams begins by simulating electrons of the primary electron beam interacting with a target. Because the electron beam characteristics of any single accelerator are unique and generally unknown, an appropriate model of an electron beam must be assumed before MC simulations can be run. The purpose of the present study is to develop a flexible framework with suitable regression models for estimating parameters of the model of primary electron beam in simulators of medical linear accelerators using real reference dose profiles measured in a water phantom. Methods All simulations were run using PRIMO MC simulator. Two regression models for estimating the parameters of the simulated primary electron beam, both based on machine learning, were developed. The first model applies Principal Component Analysis to measured dose profiles in order to extract principal features of the shapes of the these profiles. The PCA-obtained features are then used by Support Vector Regressors to estimate the parameters of the model of the electron beam. The second model, based on deep learning, consists of a set of encoders processing measured dose profiles, followed by a sequence of fully connected layers acting together, which solve the regression problem of estimating values of the electron beam parameters directly from the measured dose profiles. Results of the regression are then used to reconstruct the dose profiles based on the PCA model. Agreement between the measured and reconstructed profiles can be further improved by an optimization procedure resulting in the final estimates of the parameters of the model of the primary electron beam. These final estimates are then used to determine dose profiles in MC simulations. Results Analysed were a set of actually measured (real) dose profiles of 6 MV beams from a real Varian 2300 C/D accelerator, a set of simulated training profiles, and a separate set of simulated testing profiles, both generated for a range of parameters of the primary electron beam of the Varian 2300 C/D PRIMO simulator. Application of the two-stage procedure based on regression followed by reconstruction-based minimization of the difference between measured (real) and reconstructed profiles resulted in achieving consistent estimates of electron beam parameters and in a very good agreement between the measured and simulated photon beam profiles. Conclusions The proposed framework is a readily applicable and customizable tool which may be applied in tuning virtual primary electron beams of Monte Carlo simulators of linear accelerators. The codes, training and test data, together with readout procedures, are freely available at the site: https://github.com/taborzbislaw/DeepBeam.

DeepBeam – A Machine Learning Framework For Tuning The Primary Electron Beam of The PRIMO Monte Carlo Software

BackgroundAny Monte Carlo simulation of dose delivery using medical accelerator-generated megavolt photon beams begins by simulating electrons of the primary electron beam interacting with a target. Because the electron beam characteristics of any single accelerator are unique and generally unknown, an appropriate model of an electron beam must be assumed before MC simulations can be run. The purpose of the present study is to develop a flexible framework with suitable regression models for estimating parameters of the model of primary electron beam in simulators of medical linear accelerators, basing on real reference dose profiles measured in a water phantom. MethodsAll simulations were run using PRIMO MC simulator. Two regression models for estimating the parameters of the simulated primary electron beam, both based on machine learning, were developed. The first model applies Principal Component Analysis to measured dose profiles in order to extract principal features of the shapes of the these profiles. The PCA-obtained features are then used by Support Vector Regressors to estimate the parameters of the model of the electron beam. The second model, based on deep learning, consists of a set of encoders processing measured dose profiles, followed by a sequence of fully connected layers acting together, which solve the regression problem of estimating values of the electron beam parameters directly from the measured dose profiles. Results of the regression are then used to reconstruct the dose profiles, basing on the PCA model. Agreement between the measured and reconstructed profiles can be further improved by an optimization procedure resulting in the final estimates of the parameters of the model of the primary electron beam. These final estimates are then used to determine dose profiles in MC simulations.ResultsAnalysed were a set of actually measured (real) dose profiles of 6 MV beams from a real Varian 2300 C/D accelerator, a set of simulated training profiles, and a separate set of simulated testing profiles, both generated for a range of parameters of the primary electron beam of the Varian 2300 C/D PRIMO simulator. Application of the two-stage procedure based on regression followed by reconstruction-based minimization of the difference between measured (real) and reconstructed profiles resulted in achieving consistent estimates of electron beam parameters and in a very good agreement between the measured and simulated photon beam profiles.ConclusionsThe proposed framework is a readily applicable and customizable tool which may be applied in tuning virtual primary electron beams of Monte Carlo simulators of linear accelerators. The codes, training and test data, together with some trained models and readout procedures, are freely available at the site: https://github.com/taborzbislaw/DeepBeam.

Assessment Of Whole Body, Skin and Eye Lens Doses of the Interventional Radiologists At Selected Hospitals in Iran

High radiation doses to the body may lead to the stochastic/deterministic effects of ionizing radiation on the critical organs as well as causing the cataract in eye lens of the clinical staff in interventional radiology. In this study, the received doses of the eyes, skin and whole body of 38 clinical staff including physicians, residents, nurses and radiotechnologists in cardiac angiography departments in three selected hospitals were assessed using personal dosemeters during two bimonthly dosimetry periods. Moreover, the correlation coefficients among the measured dose components including eye lens dose, skin dose and whole body dose equivalent in both area of under and over their lead-apron were calculated for all these occupational groups. The results show that the occupational annual dose values of the clinical staff are below the annual dose limits recommended by International Commission on Radiation Protection. Furthermore, among the measured dose components, the highest correlation coefficient value was obtained between the eye lens dose and personal dose equivalent measured over the lead apron for all the occupational groups.

Dose Estimates of Occupational Radiation Exposure During Radioguided Surgery of [99mTc]Tc-PSMA-labeled Lymph Nodes in Recurrent Prostate Cancer

Background and objective[99mTc]Tc-PSMA-based radioguided surgery (TPRS) represents a curative approach for localized relapse of prostate cancer. For its simplified regulatory permission, the radiation protection authorities require a 99mTc activity below the exemption limit of 10 MBq at the time of surgery. Our aim was to determine the optimal amount of radioactivity (OAR) to comply with that limit and to estimate the maximum number of TPRS procedures per year and surgeon without triggering the full monitoring obligations.MethodsIn this retrospective study, a dose rate meter was calibrated using measurements on phantoms and from recently injected (1 min p.i.) patients to determine the activity in the patient from measured dose rates. The effective half-life of [99mTc]Tc-PSMA-I&S in patients was determined from repeated dose rate measurements up to 27 h p.i. to estimate dose parameters of relevance for radiation protection. External exposures of the surgeons were measured with personal dosimeters calibrated in Hp(10). ResultsFrom the first 6 subsequent patients, an effective half-life of 4.15 h was observed. Assuming an operation time 24 h p.i., the OAR was 550 MBq. Operations lasting in average 2 h in a distance of 0.25 m to the patient imply a body dose for surgeons of 4,16 µSv per procedure. Based on these estimates, the surgeon’s Hp(10) is less than 1 mSv per year with up to 241 operations per year. The effective dose for surgeons during the procedure determined with an electronic dosimeter is 4±1 µSv. SummaryAll radiation protection regulations are met with adherence to OAR recommended here without triggering the full monitoring obligations from radiation protection regulations.

Evaluation of Platinum-catalyzed Silicones for Fabrication of Biocompatible Patient-specific Elastic Bolus

Purpose We investigated the properties of platinum-catalyzed silicones with suitable characteristics for a biocompatible patient-specific elastic bolus. Materials & Methods We applied a platinum-catalyzed silicone (Ecoflex™ 0030) and a platinum cure liquid silicone (Dragon Skin™ 10 MEDIUM) to fabricate a biocompatible bolus using a mold and casting method with a 3D printer. We conducted physical evaluations including the shore hardness, cure time, transparency, and mixed viscosity. The dosimetric characteristics were basically investigated with surface dose and beam quality. For dosimetric evaluations using humanoid phantom, the dose differences between calculated dose and measured dose were compared with those of Dragon skin. To evaluate which boluses fit best, the volume of unwanted air gaps between the bolus and phantom was obtained from CT images. Biological evaluations were conducted on the skin sensitization, skin irritation, and cytotoxicity to ensure safe patient application. Results Ecoflex were biologically evaluated as safety materials. In addition, Ecoflex shows excellent physical properties with respect to a low shore hardness (00–30), short curing time (4 h), and low mixed viscosity (3,000). For the dosimetric properties using humanoid phantom, the average dose difference between the calculated dose and measured dose for Ecoflex is about 0.5% better than that of Dragon skin. In addition, a relatively smaller volume of unwanted air gaps for Ecoflex also showed than for Dragon skin, which these results tended to be the same as the dosimetric results. Conclusion The physical properties of Ecoflex including excellent adhesive strength, lower shore hardness reduces unwanted air gaps and ensures an accurate dose distribution. Therefore, it is a suitable material for fabricating biocompatible patient-specific elastic bolus. It would be an alternative to other materials of bolus and thus improve the efficiency for clinical use.