Evaluation of radiation detectors for the determination of field output factors in Leksell Gamma Knife dosimetry using 3D printed phantom inserts.

The Leksell Gamma Knife® Perfexion™ and Icon™ have a unique geometry, containing 192 60Co sources with collimation for field sizes of 4 mm, 8 mm, and 16 mm. 4 mm and 8 mm collimated fields lack lateral charged particle equilibrium, so accurate field output factors are essential. This study performs field output factor measurements for the microDiamond, microSilicon, and RAZOR™ Nano detectors. 3D printed inserts for the spherical Solid Water® Phantom were fabricated for microDiamond detector, the microSilicon unshielded diode and the RAZOR™ Nano micro-ionisation chamber. Detectors were moved iteratively to identify the peak detector signal for each collimator, representing the effective point of measurement of the chamber. In addition, field output correction factors were calculated for each detector relative to vendor supplied Monte Carlo simulated field output factors and field output factors measured with a W2 scintillator. All field output factors where within 1.1 % for the 4 mm collimator and within 2.3 % for the 8 mm collimator. The 3D printed phantom inserts were suitable for routine measurements if the user identifies the effective point of measurement, and ensures a reproducible setup by marking the rotational alignment of the cylindrical print. Measurements with the microDiamond and microSilicon can be performed faster compared to the RAZOR™ Nano due to differences in the signal to noise ratio. All detectors are suitable for field output factor measurements for the Leksell Gamma Knife® Perfexion™ and Icon™.

Utility of 68Ga-DOTATATE PET-MRI for Gamma Knife® stereotactic radiosurgery treatment planning for meningioma.

OBJECTIVES: To investigate the impact of adding 68Ga-DOTATATE PET/MRI to standard MRI for target volume delineation in Gamma Knife® stereotactic radiosurgery (GKSRS) for meningioma. METHODS: Seventeen patients with 18 lesions undergoing GKSRS for WHO grade 1 meningioma were enrolled in a prospective study. All patients underwent pre-treatment 68Ga-DOTATATE PET/MRI examination in addition to standard procedures. Five clinicians independently contoured the gross tumour volume (GTV) based on standard MRI (GTVMRI) and PET/MRI (GTVPET/MRI) on separate occasions. Interobserver agreement was evaluated using Cohen's Kappa statistic (CKS), Dice similarity coefficient (DC), and Hausdorff distance (HD). Statistical analysis was performed with paired t-test and Wilcoxon signed rank test. RESULTS: The addition of PET/MRI significantly increased GTV contour volume (mean GTVPET/MRI 3.59 cm3 versus mean GTVMRI 3.18 cm3, P = .008). Using the treating clinician's pre-treatment GTVMRI as the reference, median CKS (87.2 vs 77.5, P = .006) and DC (87.2 vs 77.4, P = .006) were significantly lower, and median HD (25.2 vs 31.0, P = .001) was significantly higher with the addition of PET/MRI. No significant difference was observed in interobserver contouring reproducibility between GTVMRI and GTVPET/MRI. CONCLUSION: The addition of 68Ga-DOTATATE PET/MRI for target volume delineation in GKSRS for meningioma is associated with an increase in GTV volume and greater interobserver variation. PET/MRI did not affect interobserver contouring reproducibility. ADVANCES IN KNOWLEDGE: This study provides novel insights into the impact of 68Ga-DOTATATE PET/MRI on GTV delineation and interobserver agreement in meningioma GKSRS, highlighting its potential for improving GKSRS treatment accuracy.

The first clinical implementation of real-time 6 degree-of-freedom image-guided radiotherapy for liver SABR patients.

PURPOSE: Multiple survey results have identified a demand for improved motion management for liver cancer IGRT. Until now, real-time IGRT for liver has been the domain of dedicated and expensive cancer radiotherapy systems. The purpose of this study was to clinically implement and characterise the performance of a novel real-time 6 degree-of-freedom (DoF) IGRT system, Kilovoltage Intrafraction Monitoring (KIM) for liver SABR patients. METHODS/MATERIALS: The KIM technology segmented gold fiducial markers in intra-fraction x-ray images as a surrogate for the liver tumour and converted the 2D segmented marker positions into a real-time 6DoF tumour position. Fifteen liver SABR patients were recruited and treated with KIM combined with external surrogate guidance at three radiotherapy centres in the TROG 17.03 LARK multi-institutional prospective clinical trial. Patients were either treated in breath-hold or in free breathing using the gating method. The KIM localisation accuracy and dosimetric accuracy achieved with KIM + external surrogate were measured and the results were compared to those with the estimated external surrogate alone. RESULTS: The KIM localisation accuracy was 0.2±0.9 mm (left-right), 0.3±0.6 mm (superior-inferior) and 1.2±0.8 mm (anterior-posterior) for translations and -0.1◦±0.8◦ (left-right), 0.6◦±1.2◦ (superior-inferior) and 0.1◦±0.9◦ (anterior-posterior) for rotations. The cumulative dose to the GTV with KIM + external surrogate was always within 5% of the plan. In 2 out of 15 patients, >5% dose error would have occurred to the GTV and an organ-at-risk with external surrogate alone. CONCLUSIONS: This work demonstrates that real-time 6DoF IGRT for liver can be implemented on standard radiotherapy systems to improve treatment accuracy and safety. The observations made during the treatments highlight the potential false assurance of using traditional external surrogates to assess tumour motion in patients and the need for ongoing improvement of IGRT technologies.

Repurposing traditional U-Net predictions for sparse SAM prompting in medical image segmentation.

Objective:Automated medical image segmentation (MIS) using deep learning has traditionally relied on models built and trained from scratch, or at least fine-tuned on a target dataset. The Segment Anything Model (SAM) by Meta challenges this paradigm by providing zero-shot generalisation capabilities. This study aims to develop and compare methods for refining traditional U-Net segmentations by repurposing them for automated SAM prompting.Approach:A 2D U-Net with EfficientNet-B4 encoder was trained using 4-fold cross-validation on an in-house brain metastases dataset. Segmentation predictions from each validation set were used for automatic sparse prompt generation via a bounding box prompting method (BBPM) and novel implementations of the point prompting method (PPM). The PPMs frequently produced poor slice predictions (PSPs) that required identification and substitution. A slice was identified as a PSP if it (1) contained multiple predicted regions per lesion or (2) possessed outlier foreground pixel counts relative to the patient's other slices. Each PSP was substituted with a corresponding initial U-Net or SAM BBPM prediction. The patients' mean volumetric dice similarity coefficient (DSC) was used to evaluate and compare the methods' performances.Main results:Relative to the initial U-Net segmentations, the BBPM improved mean patient DSC by 3.93 ± 1.48% to 0.847 ± 0.008 DSC. PSPs constituted 20.01-21.63% of PPMs' predictions and without substitution performance dropped by 82.94 ± 3.17% to 0.139 ± 0.023 DSC. Pairing the two PSP identification techniques yielded a sensitivity to PSPs of 92.95 ± 1.20%. By combining this approach with BBPM prediction substitution, the PPMs achieved segmentation accuracies on par with the BBPM, improving mean patient DSC by up to 4.17 ± 1.40% and reaching 0.849 ± 0.007 DSC.Significance:The proposed PSP identification and substitution techniques bridge the gap between PPM and BBPM performance for MIS. Additionally, the uniformity observed in our experiments' results demonstrates the robustness of SAM to variations in prompting style. These findings can assist in the design of both automatically and manually prompted pipelines.

Assessment of Optimizers and their Performance in Autosegmenting Lung Tumors.

Purpose: Optimizers are widely utilized across various domains to enhance desired outcomes by either maximizing or minimizing objective functions. In the context of deep learning, they help to minimize the loss function and improve model's performance. This study aims to evaluate the accuracy of different optimizers employed for autosegmentation of non-small cell lung cancer (NSCLC) target volumes on thoracic computed tomography images utilized in oncology. Materials and Methods: The study utilized 112 patients, comprising 92 patients from "The Cancer Imaging Archive" (TCIA) and 20 of our local clinical patients, to evaluate the efficacy of various optimizers. The gross tumor volume was selected as the foreground mask for training and testing the models. Of the 92 TCIA patients, 57 were used for training and validation, and the remaining 35 for testing using nnU-Net. The performance of the final model was further evaluated on the 20 local clinical patient datasets. Six different optimizers, namely AdaDelta, AdaGrad, Adam, NAdam, RMSprop, and stochastic gradient descent (SGD), were investigated. To assess the agreement between the predicted volume and the ground truth, several metrics including Dice similarity coefficient (DSC), Jaccard index, sensitivity, precision, Hausdorff distance (HD), 95th percentile Hausdorff distance (HD95), and average symmetric surface distance (ASSD) were utilized. Results: The DSC values for AdaDelta, AdaGrad, Adam, NAdam, RMSprop, and SGD were 0.75, 0.84, 0.85, 0.84, 0.83, and 0.81, respectively, for the TCIA test data. However, when the model trained on TCIA datasets was applied to the clinical datasets, the DSC, HD, HD95, and ASSD metrics showed a statistically significant decrease in performance compared to the TCIA test datasets, indicating the presence of image and/or mask heterogeneity between the data sources. Conclusion: The choice of optimizer in deep learning is a critical factor that can significantly impact the performance of autosegmentation models. However, it is worth noting that the behavior of optimizers may vary when applied to new clinical datasets, which can lead to changes in models' performance. Therefore, selecting the appropriate optimizer for a specific task is essential to ensure optimal performance and generalizability of the model to different datasets.

Auto-segmentation of thoracic organs in CT scans of breast cancer patients using a 3D U-net cascaded into 2D patchGANs.

The quality of organ volume delineation significantly influences the efficacy of radiotherapy treatment for breast cancer patients. This study introduces a novel method for auto-segmentation of the breasts, lungs and heart. The proposed pipeline leverages a multi-class 3D U-Net with a pre-trained ResNet(2+1)D-18 encoder branch, cascaded with a 2D PatchGAN mask correction model for each class. This approach requires a single 3D model, providing a relatively efficient solution. The models were trained and evaluated on 70 thoracic DICOM datasets belonging to breast cancer patients. The evaluation demonstrated state-of-the-art segmentation performance, with mean Dice similarity coefficient values ranging from 0.89 to 0.98, Hausdorff distance values ranging from 2.25 to 8.68 mm, and mean surface distance values ranging from 0.62 to 2.79 mm. These results underscore the pipeline's potential to enhance breast cancer diagnosis and treatment strategies, with possible applications in other medical sectors utilizing auto-segmentation.

Generating missing patient anatomy from partially acquired cone-beam computed tomography images using deep learning: a proof of concept.

The patient setup technique currently in practice in most radiotherapy departments utilises on-couch cone-beam computed tomography (CBCT) imaging. Patients are positioned on the treatment couch using visual markers, followed by fine adjustments to the treatment couch position depending on the shift observed between the computed tomography (CT) image acquired for treatment planning and the CBCT image acquired immediately before commencing treatment. The field of view of CBCT images is limited to the size of the kV imager which leads to the acquisition of partial CBCT scans for lateralised tumors. The cone-beam geometry results in high amounts of streaking artifacts and in conjunction with limited anatomical information reduces the registration accuracy between planning CT and the CBCT image. This study proposes a methodology that can improve radiotherapy patient setup CBCT images by removing streaking artifacts and generating the missing patient anatomy with patient-specific precision. This research was split into two separate studies. In Study A, synthetic CBCT (sCBCT) data was created and used to train two machine learning models, one for removing streaking artifacts and the other for generating the missing patient anatomy. In Study B, planning CT and on-couch CBCT data from several patients was used to train a base model, from which a transfer of learning was performed using imagery from a single patient, producing a patient-specific model. The models developed for Study A performed well at removing streaking artifacts and generating the missing anatomy. The outputs yielded in Study B show that the model understands the individual patient and can generate the missing anatomy from partial CBCT datasets. The outputs generated demonstrate that there is utility in the proposed methodology which could improve the patient setup and ultimately lead to improving overall treatment quality.

Autosegmentation of lung computed tomography datasets using deep learning U-Net architecture.

Aim: Current radiotherapy treatment techniques require a large amount of imaging data for treatment planning which demand significant clinician's time to segment target volume and organs at risk (OARs). In this study, we propose to use U-net-based architecture to segment OARs commonly encountered in lung cancer radiotherapy. Materials and Methods: Four U-Net OAR models were generated and trained on 20 lung cancer patients' computed tomography (CT) datasets, with each trained for 100 epochs. The model was tested for each OAR, including the right lung, left lung, heart, and spinal cord. Dice similarity coefficient (DSC) and Hausdorff distance (HD) were used to assess the agreement between the predicted contour and ground truth. Results: The highest of the average DSC among the test patients for the left lung and the right lung was 0.96 ± 0.03 and 0.94 ± 0.06, respectively, and 0.88 ± 0.04 for heart, and 0.76 ± 0.07 for the spinal cord. The HD for these corresponding DSCs was 3.51 ± 0.85, 4.06 ± 1.12, 4.09 ± 0.85, and 2.76 ± 0.52 mm for left lung, right lung, heart, and spinal cord, respectively. Conclusion: The autosegmented regions predicted by right and left lung models matched well with the manual contours. However, in a few cases, the heart model struggled to outline the boundary precisely. The spinal cord model had the lowest DSC, which may be due to its small size. This is an ongoing study aimed to assist radiation oncologists in segmenting the OARs with minimal effort.

Investigation of Computed Tomography Numbers on Multiple Imaging Systems using Single and Multislice Methods.

Aim: The aim of this study is to determine the variation in Hounsfield values with single and multi-slice methods using in-house software on fan-beam computed tomography (FCT), linear accelerator (linac) cone-beam computed tomography (CBCT), and Icon-CBCT datasets acquired using Gammex and advanced electron density (AED) phantoms. Materials and Methods: The AED phantom was scanned on a Toshiba computed tomography (CT) scanner, five linac-based CBCT X-ray volumetric imaging systems, and Leksell Gamma Knife Icon. The variation between single and multi-slice methods was assessed by comparing scans acquired using Gammex and AED phantoms. The variation in Hounsfield units (HUs) between seven different clinical protocols was assessed using the AED phantom. A CIRS Model 605 Radiosurgery Head Phantom (TED) phantom was scanned on all three imaging systems to assess the target dosimetric changes due to HU variation. An in-house software was developed in MATLAB to assess the HU statistics and the trend along the longitudinal axis. Results: The FCT dataset showed a minimal variation (central slice ± 3 HU) in HU values along the long axis. A similar trend was also observed between the studied clinical protocols acquired on FCT. Variation among multiple linac CBCTs was insignificant. In the case of the water insert, a maximum HU variation of -7.23 ± 68.67 was observed for Linac 1 towards the inferior end of the phantom. All five linacs appeared to have a similar trend in terms of HU variation from the proximal to the distal end of the phantom, with a few outliers for Linac 5. Among three imaging modalities, the maximum variation was observed in gamma knife CBCTs, whereas FCT showed no appreciable deviation from the central value. In terms of dosimetric comparison, the mean dose in CT and Linac CBCT scans differed by <0.5 Gy, whereas at least a 1 Gy difference was observed between CT and gamma knife CBCT. Conclusion: This study shows a minimal variation with FCT between single, volume-based, and multislice methods, and hence the current approach of determining the CT-electron density curve based on a single-slice method would be sufficient for producing a HU calibrations curve for treatment planning. However, CBCTs acquired on linac, and in particular, gamma knife systems, show noticeable variations along the long axis, which is likely to affect the dose calculations performed on CBCTs. It is highly recommended to assess the Hounsfield values on multiple slices before using the HU curve for dose calculations.

Design and Development of a Low-cost Integrated Dosimeter for External Beam Dosimetry in Radiation Oncology.

Radiation dosimeters play a crucial role in radiation oncology by accurately measuring radiation dose, ensuring precise and safe radiation therapy. This study presents the design and development of a low-cost printed circuit board (PCB) dosimeter and an integrated electrometer with sensitivity optimized for dose rates intended for use in megavoltage radiation therapy. The PCB dosimeter was designed in KiCad, and it uses a low-cost S5MC-13F general-purpose 1 kV 5A power diode as a radiation detector. The dosimeter is calibrated against a known dose derived from an ionization chamber and tested for dose linearity, dose rate dependence, field size dependence, and detector orientation dependence. The observed average dose differences between the delivered and measured doses for most measurements were found to be < 1.1%; the dose rate linearity between 100 MU/min and 1400 MU/min was found to be within 1.3%. This low-cost architecture could successfully be adapted further for a scalable, cost-effective dosimetry solution through firmware or circuit design.

Autosegmentation of brain metastases using 3D FCNN models and methods to manage GPU memory limitations.

Aims. To explore the efficacy of two different approaches to train a Fully Convolutional Neural Network (FCNN) with Graphical Processing Unit (GPU) memory limitations and investigate if pre-trained two-dimensional weights can be transferred into a three-dimensional model for the purpose of brain tumour segmentation.Materials and methods. Models were developed in Python using TensorFlow and Keras. T1 contrast-enhanced MRI scans and associated contouring data from 104 patients were used to train and validate the model. The data was resized to one-quarter of its original resolution, and the original data was also split into four quarters for comparison to fit within GPU limitations. Transferred weights from a two-dimensional VGG16 model trained on ImageNet were transformed into three dimensions for comparison with randomly generated initial weights.Results. Resizing the data produced superior dice similarity coefficients with fewer false positives than quartering the data. Quartering the data yielded a superior sensitivity. Transforming and transferring two-dimensional weights was not able to consistently produce improvement in training or final metrics.Conclusion. For segmentation of brain tumours, resizing the data results in better performance than quartering the data. For the model and approaches used in this report, transferring weights were not able to demonstrate any benefit.

A direct comparison of the optically stimulated luminescent properties of BeO and Al2O3 for clinical in-vivo dosimetry.

Optically stimulated luminescence dosimetry is a relatively recent field of in-vivo dosimetry in clinical radiotherapy, developing over the last 20 years. As a pilot study, this paper presents a direct comparison between the sensitivity variance with use, stability of measurement and linearity of the current clinical standard Al2O3:C and a potential alternative, beryllium oxide. A set of ten optically stimulated luminescence dosimeters (OSLD), including five of each type, were used simultaneously and irradiated on a Versa HD linear accelerator. Having similar sensitivity, while Al2O3:C showed a relatively stable signal response from initial use, BeO was found to have a higher response to the same dose. However, BeO displayed a strong exponential decline from initial signal response following a model of [Formula: see text], reaching stability after approximately 10 irradiation cycles. BeO was shown to have potentially higher accuracy than Al2O3:C, with less variation between individual doses. Both OSLD showed good linearity between 0.2-5.0 Gy. Between these bounds, Al2O3:C demonstrated a strong linear response following the trend [Formula: see text], however beyond this showed deviation from linearity, resulting in a measured dose of [Formula: see text] Gy at 10.0 Gy dose delivery. BeO showed strong linearity across the full examined range of 0.2-10.0 Gy with following a model of [Formula: see text] Gy with a recorded dose at 10.0 Gy delivery as [Formula: see text] Gy. In conclusion, BeO does show large variance in sensitivity between individual OSLD and a considerable initial variance and decline in dose-response, however after pre-conditioning and individual normalisation to offset OSLD specific sensitivity BeO provides not only a viable alternative to Al2O3:C, but potentially provide higher accuracy, precision and reproducibility for in-vivo dosimetry.

Dose assessment for daily cone-beam CT in lung radiotherapy patients and its combination with treatment planning.

With the increased use of X-ray imaging for patient alignment in external beam radiation therapy, particularly with cone-beam computed tomography (CBCT), the additional dose received by patients has become of greater consideration. In this study, we analysed the radiation dose from CBCT for clinical lung radiotherapy and assessed its relative contribution when combined with radiation treatment planning for a variety of lung radiotherapy techniques. The Monte Carlo simulation program ImpactMC was used to calculate the 3D dose delivered by a Varian TrueBeam linear accelerator to patients undergoing thorax CBCT imaging. The concomitant dose was calculated by simulating the daily CBCT irradiation of ten lung cancer patients. Each case was planned with a total dose of 50-60 Gy to the target lesion in 25-30 fractions using the 3DCRT or IMRT plan and retrospectively planned using VMAT. For each clinical case, the calculated CBCT dose was summed with the planned dose, and the dose to lungs, heart, and spinal cord were analysed according to conventional dose conformity metrics. Our results indicate greater variations in dose to the heart, lungs, and spinal cord based on planning technique, (3DCRT, IMRT, VMAT) than from the inclusion of daily cone-beam imaging doses over 25-30 fractions. The average doses from CBCT imaging per fraction to the lungs, heart and spinal cord were 0.52 ± 0.10, 0.49 ± 0.15 and 0.39 ± 0.08 cGy, respectively. Lung dose variations were related to the patient's size and body composition. Over a treatment course, this may result in an additional mean absorbed dose of 0.15-0.2 Gy. For lung V5, the imaging dose resulted in an average increase of ~ 0.6% of the total volume receiving 5 Gy. The increase in V20 was more dependent on the planning technique, with 3DCRT increasing by 0.11 ± 0.09% with imaging and IMRT and VMAT increasing by 0.17 ± 0.05% and 0.2 ± 0.06%, respectively. In this study, we assessed the concomitant dose for daily CBCT lung cancer patients undergoing radiotherapy. The additional radiation dose to the normal lungs from daily CBCT was found to range from 0.15 to 0.2 Gy when the patient was treated with 25-30 fractions. Consideration of potential variation in relative biological effectiveness between kilovoltage imaging and megavoltage treatment dose was outside the scope of this study. Regardless of this, our results show that the assessment of imaging dose can be incorporated into the treatment planning process and the relative effect on overall dose distribution was small compared to the difference among planning techniques.

A feasibility study on the development and use of a deep learning model to automate real-time monitoring of tumor position and assessment of interfraction fiducial marker migration in prostate radiotherapy patients.

Purpose. The aim of this study was to assess the feasibility of the development and training of a deep learning object detection model for automating the assessment of fiducial marker migration and tracking of the prostate in radiotherapy patients.Methods and Materials. A fiducial marker detection model was trained on the YOLO v2 detection framework using approximately 20,000 pelvis kV projection images with fiducial markers labelled. The ability of the trained model to detect marker positions was validated by tracking the motion of markers in a respiratory phantom and comparing detection data with the expected displacement from a reference position. Marker migration was then assessed in 14 prostate radiotherapy patients using the detector for comparison with previously conducted studies. This was done by determining variations in intermarker distance between the first and subsequent fractions in each patient.Results. On completion of training, a detection model was developed that operated at a 96% detection efficacy and with a root mean square error of 0.3 pixels. By determining the displacement from a reference position in a respiratory phantom, experimentally and with the detector it was found that the detector was able to compute displacements with a mean accuracy of 97.8% when compared to the actual values. Interfraction marker migration was measured in 14 patients and the average and maximum±standard deviation marker migration were found to be2.0±0.9mmand2.3±0.9mm,respectively.Conclusion. This study demonstrates the benefits of pairing deep learning object detection, and image-guided radiotherapy and how a workflow to automate the assessment of organ motion and seed migration during prostate radiotherapy can be developed. The high detection efficacy and low error make evident the advantages of using a pre-trained model to automate the assessment of the target volume positional variation and the migration of fiducial markers between fractions.

A Framework for ExacTrac Dynamic Commissioning for Stereotactic Radiosurgery and Stereotactic Ablative Radiotherapy.

This paper aims to provide guidance and a framework for commissioning tests and tolerances for the ExacTrac Dynamic image-guided and surface-guided radiotherapy (SGRT) system. ExacTrac Dynamic includes a stereoscopic X-ray system, a structured light projector, stereoscopic cameras, thermal camera for SGRT, and has the capability to track breath holds and internal markers. The system provides fast and accurate image guidance and intrafraction guidance for stereotactic radiosurgery and stereotactic ablative radiotherapy. ExacTrac Dynamic was commissioned on a recently installed Elekta Versa HD. Commissioning tests are described including safety, isocenter calibration, dosimetry, image quality, data transfer, SGRT stability, SGRT localization, gating, fusion, implanted markers, breath hold, and end-to-end testing. Custom phantom designs have been implemented for assessment of the deep inspiration breath-hold workflow, the implanted markers workflow, and for gating tests where remote-controlled movement of a phantom is required. Commissioning tests were all found to be in tolerance, with maximum translational and rotational deviations in SGRT of 0.3 mm and 0.4°, respectively, and X-ray image fusion reproducibility standard deviation of 0.08 mm. Tolerances were based on published documents and upon the performance characteristics of the system as specified by the vendor. The unique configuration of ExacTrac Dynamic requires the end user to design commissioning tests that validate the system for use in the clinical implementation adopted in the department. As there are multiple customizable workflows available, tests should be designed around these workflows, and can be ongoing as workflows are progressively introduced into departmental procedures.

Comparison and Validation of Multiple Detectors against Monte Carlo Simulation for the Use of Small-Field Dosimetry.

Aim: The aim of this study was to compare the Exradin W2 scintillator, PTW microDiamond, IBA Razor Nano, and IBA Razor chamber detectors for small-field dose measurements and validate the measured data against the EGSnrc user code and observe the variation between daisy-chained and direct measurement methods for the above detectors. Materials and Methods: The W2 scintillator, microDiamond, Razor Nano, and Razor chamber detectors were used to measure the in-plane and cross-plane profiles and the output factors (OFs) at 10 cm depth, and 90 source-to-surface distance for 6MV X-rays (Elekta Versa HD). The field sizes ranged from 0.5 cm × 0.5 cm to 5 cm × 5 cm. The BEAMnrc/DOSXYZnrc user codes (EGSnrc) were used to simulate the reference profiles. Gamma analysis was performed to compare the measured and simulated dose distributions. Results: The OFs measured with the W2 scintillator, microDiamond, Razor Nano chamber, Razor chamber, and the calculated Monte Carlo (MC) showed agreement to within 1% for the 3 cm × 3 cm field size. The uncertainty in the MC simulation was observed to be 0.4%. The percent difference in OFs measured using daisy-chained and direct measurement methods was within 0.15%, 0.4%, 1.4%, and 2.4% for microDiamond, W2 scintillator, Nano, and Razor chamber detectors, respectively. Conclusion: The lateral beam profiles and OFs of W2 scintillator, microDiamond, Razor Nano, and Razor chambers exhibit good agreement with the MC simulation within the nominal field sizes. Our results demonstrate that we can achieve considerable time-saving by directly measuring small-field OFs without daisy-chained methods using microDiamond and W2 scintillator. In terms of ease of use, sensitivity, reproducibility, and from a practical standpoint, we recommend microDiamond for small-field dosimetry.

Use of GammaPlan convolution algorithm for dose calculation on CT and cone-beam CT images.

PURPOSE: The aim of this study was to assess the suitability of using cone-beam computed tomography images (CBCTs) produced in a Leksell Gamma Knife (LGK) Icon system to generate electron density information for the convolution algorithm in Leksell GammaPlan (LGP) Treatment Planning System (TPS). MATERIALS AND METHODS: A retrospective set of 30 LGK treatment plans generated for patients with multiple metastases was selected in this study. Both CBCTs and fan-beam CTs were used to provide electron density data for the convolution algorithm. Plan quality metrics such as coverage, selectivity, gradient index, and beam-on time were used to assess the changes introduced by convolution using CBCT (convCBCT) and planning CT (convCT) data compared to the homogeneous TMR10 algorithm. RESULTS: The mean beam-on time for TMR10 and convCBCT was found to be 18.9 ± 5.8 minutes and 21.7 ± 6.6 minutes, respectively. The absolute mean difference between TMR10 and convCBCT for coverage, selectivity, and gradient index were 0.001, 0.02, and 0.0002, respectively. The calculated beam-on times for convCBCT were higher than the time calculated for convCT treatment plans. This is attributed to the considerable variation in Hounsfield values (HU) dependent on the position within the field of view. CONCLUSION: The artifacts from the CBCT's limited field-of-view and considerable HU variation need to be taken into account before considering the use of convolution algorithm for dose calculation on CBCT image datasets, and electron data derived from the onboard CBCT should be used with caution.

SABR pre-treatment checks using alanine and nanoDot dosimeters.

Stereotactic Ablative Radiotherapy (SABR) remains one of the preferred treatment techniques for early-stage cancer. It can be extended to more treatment locales involving the sternum, scapula and spine. This work investigates SABR checks using Alanine and nanoDot dosimeter for three treatment sites, including sternum, spine and scapula. Alanine and nanoDot dosimeters' performances were verified using a 6 MV photon beam before SABR pretreatment verifications. Each dosimeter was placed inside customized designed inserts into a Rod Phantom (in-house phantom) made of Perspex that mimics the human body for a SABR check. Electron Paramagnetic Resonance (EPR) spectrometer, Bruker EleXsys E500 (9.5 GHz) and Microstar (Landauer Inc.) Reader was employed to acquire the irradiated alanine and nanoDot dosimeters' signal, respectively. Both dosimeters treatment sites are expressed as mean ± standard deviation (SD) of the measured and Eclipse calculated dose Alanine (19.59 ± 0.24, 17.98 ± 0.15, 17.95 ± 0.18) and nanoDot (19.70 ± 0.43, 17.05 ± 0.08, 17.95 ± 0.98) for spine, scapula and sternum, respectively. The percentage difference between alanine and nanoDot dosimeters was within 2% for sternum and scapula but 2.4% for spine cases. These results demonstrate Alanine and nanoDot dosimeters' potential usefulness for SABR pretreatment quality assurance (QA).

Automatic Detection and Tracking of Marker Seeds Implanted in Prostate Cancer Patients using a Deep Learning Algorithm.

PURPOSE: Fiducial marker seeds are often used as a surrogate to identify and track the positioning of prostate volume in the treatment of prostate cancer. Tracking the movement of prostate seeds aids in minimizing the prescription dose spillage outside the target volume to reduce normal tissue complications. In this study, You Only Look Once (YOLO) v2™ (MathWorks™) convolutional neural network was employed to train ground truth datasets and develop a program in MATLAB that can visualize and detect the seeds on projection images obtained from kilovoltage (kV) X-ray volume imaging (XVI) panel (Elekta™). METHODS: As a proof of concept, a wax phantom containing three gold marker seeds was imaged, and kV XVI seed images were labeled and used as ground truth to train the model. The projection images were corrected for any panel shift using flex map data. Upon successful testing, labeled marker seeds and projection images of three patients were used to train a model to detect fiducial marker seeds. A software program was developed to display the projection images in real-time and predict the seeds using YOLO v2 and determine the centers of the marker seeds on each image. RESULTS: The fiducial marker seeds were successfully detected in 98% of images from all gantry angles; the variation in the position of the seed center was within ± 1 mm. The percentage difference between the ground truth and the detected seeds was within 3%. CONCLUSION: Our study shows that deep learning can be used to detect fiducial marker seeds in kV images in real time. This is an ongoing study, and work is underway to extend it to other sites for tracking moving structures with minimal effort.

Contralateral breast dose with electronic compensators and conventional tangential fields - A clinical dosimetric study.

Dose to the contralateral breast (CLB) from radiotherapy treatment has the potential to induce secondary breast cancer. Electronic tissue compensation (eComp) for breast cancer patients is one of the alternative methods to conventional 3D-conformal radiotherapy that eliminates the use of wedges. Several studies have investigated dose to the CLB using tangential fields involving wedges, intensity-modulated radiation therapy and volumetric modulated arc radiation therapy and various other techniques via treatment planning system calculations, Monte Carlo methods and phantoms. However, there are limited data published in assessing the actual dose received by the CLB from treatment with eComp-based tangential fields. In this study, the CLB dose for patients undergoing tangential field radiotherapy with eComp and enhanced dynamic wedged (standard) tangential fields was measured and compared to assess the CLB dose between the two methods. Measurements were conducted on a randomised trial of 40 patients, 20 of them had undergone standard planning, and the remaining 20 were treated with eComp. The mean surface dose measured with TLDs at 3cm from the medial tangential border for eComp and standard techniques was 10.04±1.37% and 10.14±2.05%, respectively for a prescription dose of 2.65Gy/fraction. The estimated dose at 1cm depth in tissue, measured with the use of perspex domes placed over the TLD at the same location, was 5.12±0.87% and 6.29±2.01% for eComp and standard, respectively. The CLB dose is dependent on the proximity of the medial tangential field edge to the contralateral breast and is patient-specific. The results of this study show that at 1cm depth, eComp technique delivers significantly less dose (p<0.05) to the CLB as compared to standard tangential fields. Furthermore, the surface dose measured for both eComp and standard are comparable indicating that the eComp-based tangential field technique does not contribute any excess dose to CLB when compared to standard tangential fields. The excess relative risk (ERR) for radiation-induced cancers for eComp was found to be 0.08, compared to 0.11 for standard tangential fields.