Dose Verification Research Articles

Synthesis of pseudo-PET/CT fusion images in radiotherapy based on a new transformer model.

PET/CT and planning CT are commonly used medical images in radiotherapy for esophageal and nasopharyngeal cancer. However, repeated scans will expose patients to additional radiation doses and also introduce registration errors. This multimodal treatment approach is expected to be further improved. A new Transformer model is proposed to obtain pseudo-PET/CT fusion images for esophageal and nasopharyngeal cancer radiotherapy. The data of 129 cases of esophageal cancer and 141 cases of nasopharyngeal cancer were retrospectively selected for training, validation, and testing. PET and CT images are used as input. Based on the Transformer model with a "focus-disperse" attention mechanism and multi-consistency loss constraints, the feature information in two images is effectively captured. This ultimately results in the synthesis of pseudo-PET/CT fusion images with enhanced tumor region imaging. During the testing phase, the accuracy of pseudo-PET/CT fusion images was verified in anatomy and dosimetry, and two prospective cases were selected for further dose verification. In terms of anatomical verification, the PET/CT fusion image obtained using the wavelet fusion algorithm was used as the ground truth image after correction by clinicians. The evaluation metrics, including peak signal-to-noise ratio, structural similarity index, mean absolute error, and normalized root mean square error, between the pseudo-fused images obtained based on the proposed model and ground truth, are represented by means (standard deviation). They are 37.82 (1.57), 95.23 (2.60), 29.70 (2.49), and 9.48 (0.32), respectively. These numerical values outperform those of the state-of-the-art deep learning comparative models. In terms of dosimetry validation, based on a 3%/2mm gamma analysis, the average passing rates of global and tumor regions between the pseudo-fused images (with a PET/CT weight ratio of 2:8) and the planning CT images are 97.2% and 95.5%, respectively. These numerical outcomes are superior to those of pseudo-PET/CT fusion images with other weight ratios. This pseudo-PET/CT fusion images obtained based on the proposed model hold promise as a new modality in the radiotherapy for esophageal and nasopharyngeal cancer.

Quantitative Evaluation of a Fully Automated Planning Solution for Prostate-Only and Whole-Pelvic Radiotherapy

Background/Objectives: To evaluate an end-to-end pipeline for normo-fractionated prostate-only and whole-pelvic cancer treatments that requires minimal human input and generates a machine-deliverable plan as an output. Methods: In collaboration with TheraPanacea, a treatment planning pipeline was developed that takes as its input a planning CT with organs-at-risk (OARs) and planning target volume (PTV) contours, the targeted linac machine, and the prescription dose. The primary components are (i) dose prediction by a single deep learning model for both localizations and (ii) a direct aperture VMAT plan optimization that seeks to mimic the predicted dose. The deep learning model was trained on 238 cases, and a held-out set of 86 cases was used for model validation. An end-to-end clinical evaluation study was performed on another 40 cases (20 prostate-only, 20 whole-pelvic). First, a quantitative evaluation was performed based on dose–volume histogram (DVH) points and plan parameter metrics. Then, the plan deliverability was assessed via portal dosimetry using the global gamma index. Additionally, the reference clinical manual plans were compared with the automated plans in terms of monitor unit (MU) numbers and modulation complexity scores (MCSv). Results: The automated plans provided adequate treatment plans (or minor deviations) with respect to the dose constraints, and the quality of the plans was similar to the manual plans for both localizations. Moreover, the automated plans showed successful deliverability and passed the portal dose verification. Despite higher median total MUs, no statistically significant correlation was observed between any of the gamma criteria tested and the number of MUs or MCSv. Conclusions: This study shows the feasibility of a deep learning-based fully automated treatment planning pipeline that generates high-quality plans that are competitive with manually made plans and are clinically approved in terms of dosimetry and machine deliverability.

A multi-institutional trial evaluating the use of an integrated quality assurance phantom for frameless single-isocenter multitarget stereotactic radiosurgery.

Brain radiosurgery treatments require multiple quality-assurance (QA) procedures to ensure accurate and precise treatment delivery of ablative doses. As single-isocenter multitarget radiosurgery treatments become more popular for treating patients with multiple brain metastases, quantifying off-axis accuracy of linear accelerators is crucial. In this study, we developed a novel brain radiosurgery integrated phantom and validated this phantom at multiple institutions to enable radiosurgery QA with a single phantom to facilitate implementation of a frameless single-isocenter, multitarget radiosurgery program. The phantom combines multiple independent verification system tests including the Winston-Lutz test, off-axis accuracy evaluation (i.e., off-axis Winston-Lutz), as well as dosimetric measurements utilizing both point dose and film measurement. A novel 3D-printed phantom, coined OneIso, was designed with a movable insert which can switch between Winston-Lutz test targets and dose measurement without moving the phantom itself. In total, four phantoms were printed, and eight institutions participated in this study, which included both Varian TrueBeam (n=6) and Elekta Versa (n=2) linear accelerators. For off-axis Winston-Lutz measurements, a row of off-axis ball-bearings (BBs) was integrated into the OneIso. To quantify the spatial accuracy versus distance from isocenter, two-dimensional displacements were calculated between the planned and delivered BB locations relative to their respective MLC-defined field borders. For dose verification, brain radiosurgery clinical treatment plans previously treated were delivered at multiple cancer centers (six of eight centers). Radiochromic film and pinpoint ion chamber comparison measurements were obtained with OneIso. Dose verification performed using the OneIso phantom across the different centers were all within on average 3% agreement, for both film and point-dose measurements. OneIso identified a reduction in spatial accuracy further away from isocenter for all eight radiosurgery machines. Differences increased as distance from isocenter increased, exceeding recommended radiosurgery accuracy tolerances (<1mm) at different distances for each machine (2-7cm), indicating that the tolerance is machine-dependent. OneIso provides a streamlined, single-setup workflow for single-isocenter multitarget frameless linac-based radiosurgery QA that can be easily translated to multiple institutions. Additionally, quantifying off-axis spatial discrepancies allows for determination of the maximum distance between targets and iso that meet single-isocenter multitarget radiosurgery program recommendations.

Gel dosimetry: An overview of dosimetry systems and read out methods

Gel dosimetry has emerged over the past three decades in response to a growing need in high-precision radiotherapy to assess, in three dimensions, the absorbed radiation dose, as would be administered in cancer patients.Radiation-induced reaction mechanisms are dependent on the class of gel dosimeter, with four classes emerging as primary dosimeters for use in radiation therapy dose verification: (i) Fricke gel dosimeters contain a Fricke solution consisting of ammonium iron (II) sulfate in an acidic solution of sulfuric acid. In Fricke systems an oxidation of ferrous ions results in a change in the nuclear magnetic resonance (NMR) relaxation rate, which enables reading out Fricke gel dosimeters by use of MRI. The radiation-induced oxidation in Fricke gel dosimeters can also be visualized by adding a redox indicator. (ii) Polymer gel dosimeters exploit the radiation induced polymerization reaction of vinyl monomers and are predominantly read out by quantitative MRI or X-ray CT. (iii) Radiochromic dosimeters do not demonstrate a significant radiation-induced change in NMR properties but can be scanned by use of optical scanners (optical CT). In contrast to Fricke gel dosimeters, radiochromic gel dosimeters do not rely on the oxidation of a metal ion but exhibit a color change upon radiation. (iv) Radiofluorogenic dosimeters become fluorescent when exposed to ionizing radiation and can be read out with a planar scanning light beam.Likewise, the imaging modality used to extract quantitative dose information depends on the class of dosimeter being used, and three primary imaging modalities have emerged in this context: quantitative MRI, x-ray CT, and optical CT imaging. The accuracy and precision of the dose information extracted from gel dosimetry systems depends on both the dosimetric properties of the gel dosimeters and the readout technique, and the optimal readout method depends on the gel dosimeter response.Despite remaining an active field of research and illustrations of the application of gel dosimetry for the validation of clinical dose distributions, the utilization of gel dosimetry as a routine clinical dosimeter has been rather limited. However, with the introduction of new radiotherapy techniques that focus on organ motion compensation, new fractionation schemes, and extreme dose rates, the need for 3D radiation dosimetry is apparent. Even with the need for 3D dosimetry being apparent, gel dosimetry faces continued challenges in areas regarding the extraction of reproducible, accurate, and precise dose information.This review paper focuses on an introduction to gel dosimeter classes; a detailed examination of the three readout techniques with emphasis on the achievable accuracy, precision, and optimization of readout parameters; an outlook on future applications in emerging new radiotherapy techniques. We note that the introduction of theragnostic hybrid MRI-Linacs that combine an MRI-scanner and a clinical linear accelerator create new opportunities for inline polymer gel dosimetry. Likewise, the use of cone beam CT on linear accelerators opens up the possibility to read out gel dosimeters on the linac. Multiple optical CT designs have shown that optical CT gel dosimetry is eminently capable of providing high quality dosimetric information from clinically relevant treatment regimes. As a result, gel dosimetry provides exciting opportunities for 3D radiation dosimetry that were not available even a few years ago. The unique feature set of a properly executed gel dosimetry workflow allows for the extraction of dosimetric information, in 3D, that is not possible with any other dosimetry system and hence gel dosimetry provides exciting opportunities for clinical and research work in the area of radiation dose measurement.

Utilizing acrylic acid polymer hydrogel for 3-D quality assurance in CyberKnife radiotherapy

Dosimetry tools play a crucial role in radiotherapy as they are essential for recording and validating intricate 3-D dose distributions. This research introduces and examines a novel acrylic acid polymer hydrogel (ACAPHG) dosimeter composition designed for 3-D dose verification and quality assurance in the context of radiotherapy treatment. A phantom made of an 80 mm diameter cylindrical glass container was utilized. The phantom contained the hydrogel, which served as the medium for radiation exposure. The water equivalent hydrogel within the phantom was subjected to irradiation by a CyberKnife robotic radiotherapy system. An optical computed tomography (OCT) scanner with sub-millimeter resolution was used to obtain imaging data. The dose distribution of a CyberKnife robotic SRS/SBRT treatment plan for a brain cancer patient was compared to that of the hydrogel's OCT scan using 2-D and 3-D gamma analysis with a criterion of 3% dose difference and 3 mm distance-to-agreement. A gamma pass rate of 94.1% for the 2-D gamma analysis and a pass rate of 99% for the 3-D gamma analysis were calculated within the region at which the treatment planning system data drops to 20% of the maximum dose. The use of the ACAPHG dosimeter in conjunction with the described setup suggests that it has the potential to offer an accurate 3-D verification of complex dose distributions in SRS/SBRT radiotherapy treatments. By employing the ACAPHG dosimeter and utilizing OCT scanning, this dosimeter enables the assessment and validation of intricate dose distributions in these advanced radiotherapy treatments.

Quality assurance for online adaptive radiotherapy: a secondary dose verification model with geometry-encoded U-Net

Abstract Background &amp; Purpose: In online adaptive radiotherapy (ART), quick computation-based secondary dose verification is crucial for ensuring the quality of ART plans while the patient is positioned on the treatment couch. However, traditional dose verification algorithms are generally time-consuming, reducing the efficiency of ART workflow. This study aims to develop an ultra-fast deep-learning (DL) based secondary dose verification algorithm to accurately estimate dose distributions using CT and fluence maps (FMs). &amp;#xD;Methods: We integrated FMs into the CT image domain by explicitly resolving the geometry of treatment delivery. For each gantry angle, an FM was constructed based on the optimized multi-leaf collimator apertures and corresponding monitoring units (MUs). To effectively encode treatment beam configuration, the constructed FMs were back-projected to 30 cm away from the isocenter with respect to the exact geometry of the treatment machines. Then, a 3D U-Net was utilized to take the integrated CT and FM volume as input to estimate dose. Training and validation were performed on 381 prostate cancer cases, with an additional 40 testing cases for independent evaluation of model performance.&amp;#xD;Results: The proposed model can estimate dose in ~15 ms for each patient. The average γ passing rate (3%/2mm, 10% threshold) for the estimated dose was 99.9% ± 0.15% on testing patients. The mean dose differences for the planning target volume (PTV) and organs at risk (OARs) were 0.07%±0.34% and 0.48%±0.72%, respectively.&amp;#xD;Conclusion: We have developed a geometry-resolved DL framework for accurate dose estimation and demonstrated its potential in real-time online ART doses verification.&amp;#xD;

Verification of mean dose in a tumor using 3D-printed tumor-model scintillators in anthropomorphic phantoms and a spherical phantom

To enhance treatment outcomes of radiotherapy, accurate verification of dosimetric parameters such as the mean, minimum, and maximum doses is essential. This study aimed to measure the mean absorbed dose in a tumor-shaped target and compare the results with the calculated value of the treatment planning system. Tumor model scintillators (TMSs) mimicking brain tumors were fabricated with a commercial 3D printer based on stereolithography files of tumors. The TMSs were irradiated following treatment plans made with the Leksell Gamma Plan® (Elekta AB, Stockholm, Sweden), and the mean dose in each TMS was measured three times and compared with the calculated value. Eleven TMSs were set at the center of a spherical solid water phantom. Two additional TMSs were fabricated according to images of vestibular schwannomas and were set at the tumor location in head model phantoms following each patient's head shape. The mean relative differences between the measured and calculated mean doses were less than one percent. This method provides an independent parameter for evaluating the accuracy of the treatment planning system.

Use the external radiotherapy StarTrack detector to study the electron beam profile.

To assess the electron beam dose verification, and to investigate the dependability and stability of the Startrack two-dimensional array. The is cross-secional study was conducted from January to April 2021 at the Baghdad Centre for Radiotherapy and Nuclear Medicine, Baghdad Medical City, Baghdad, Iraq. Quality assurance measurements were made using a StarTrack two-dimensional detector on an electron beam with an energy of 12MeV at 1.5cm depth for field sizes 6cm×6cm and 14cm×14cm. The detector was positioned 100cm from the source to the surface of the linear accelerator infinity. Testing was done using the International Electrotechnical Commissioning protocol. There was a significant difference between field sizes 6cmx6cm and 14cmx14cm (-4.97±0.83, -2.01±0.68, respectively) (p<0.05). The beam profile measurements of penumbra, flatness and symmetry were within limits ±2 cm, and ±2 cm, 2%, respectively. The percentage of output dose curve against distance showed instability in crossline for 6cm x 6cm and 14cm x 14cm field size. The output dose was found to be above the tolerance level, and should be corrected before patient treatment for 12MeV energy electron beam therapy.

Optically stimulated luminescence and radiophotoluminescence in NaMgF3:Eu nanoparticles

Optically stimulated luminescence and radiophotoluminescence were observed in NaMgF3:0.1%Eu nanoparticles synthesised via a hydrothermal technique. X-ray irradiation created F-centres, and subsequent optical stimulation at 420 nm released the electrons and electron-hole recombination produced Eu2+ emissions. The integrated optically stimulated luminescence intensity was linear up to 180 Gy. Photoluminescence was observed from Eu2+ and Eu3+ valences, whereas optically stimulated luminescence only occurred via the Eu2+ sites. X-ray irradiation also resulted in Eu3+ → Eu2+ and O2− → O− radiophotoluminescence effects, where the luminescence intensity ratio, RPLEu2+/RPLO2−, indicated a large detection range (>1 kGy). The observation of optically stimulated luminescence in lanthanide-doped near tissue-equivalent NaMgF3 nanoparticles has potential applications including as high spatial resolution two- and three-dimensional dosimeters for radiotherapy dose verification and validation. Radiophotoluminescence from oxygen-related centres and Eu is advantageous because it can provide a record of the dose history as well as an additional measure of the dose.

Feasibility of the Compton Suppressed Gamma-ray spectrometry for proton therapy applications

The presence of the Bragg peak and its use in proton therapy cause less damage to healthy tissues. For optimal use of the proton beam in proton therapy, it is necessary to monitor the deviation of its range during treatment. The prompt gamma rays originated from the nuclear reactions of the proton beam with tissue elements can be applied for this purpose. The spatial distribution of prompt gamma rays can be used to estimate the range of the incident charged particles. Detecting a wide energy range of gamma rays, up to 10 MeV, is the most challenging step of the proton dose verification by using prompt gamma spectroscopy. Compton suppression is a system that is used to reduce the contribution of scattered gamma rays. This technique will cause better detection of the peaks of low-energy prompt gamma rays. In this research, the performance of a Compton suppression system has been investigated in two fields for proton therapy and for elemental analysis of materials using isotopes by a wide range of gamma energies. At first, by using the simulated Compton suppression system, the calculations of the prompt gamma ray spectra emitted from the brain phantom during the radiation of 150 MeV protons were presented. The results indicated that the low energy elements peaks such as 40Ca (1.37 MeV) with CSF equal to 1.138 appeared in the suppressed spectrum. Also, it was possible to detect 14N (2.31 MeV) and 16O (6.13 MeV) peaks equivalent to the CSF 2.449 and 2.067 respectively. In the next step, it was possible to detect the low energy peaks of 152Eu and 155Eu isotopes in the suppressed spectrum. For this purpose, the minimum relative intensity required to detect all 152Eu and 155Eu peaks was equivalent to 0.001 and 0.002 respectively. CSF for the lowest energies such as 105 keV and 344 keV was equal to 1.169 and 1.184. The simulation was done using the GEANT4 Monte-Carlo toolkit.

Visual Dose Monitoring for Whole Breast Radiation Therapy Treatments via Combined Cherenkov Imaging and Scintillation Dosimetry

PurposeThis study investigates scintillation dosimetry coupled with Cherenkov imaging for in vivo dose monitoring during whole breast radiotherapy (WBRT). Given recent observations of excess dose to the contralateral breast (CB), in vivo dosimetry (IVD) could help ensure accurate dose delivery and decrease risks of secondary cancer. This work presents a rapid, streamlined alternative to traditional IVD, providing direct visualization of measurement location relative to the treatment field on the patient. Methods and Materials10 WBRT patients consented under an IRB-approved protocol were monitored with scintillation dosimetry and always-on Cherenkov imaging, on both their treated and CB for one to three fractions. Scintillator dosimeters, small plastic discs 1mm thick and 15mm in diameter, were calibrated against optically-stimulated luminescent dosimeters (OSLDs) to generate an integral output-to-dose conversion, where integral output is measured in post-processing through a custom fitting algorithm. The discs have been extensively characterized in a previous study for various treatment conditions including beam energy and treatment geometry. ResultsN=44 dosimetry measurements were evaluated, including 22 treated breast and 22 CB measurements. Following integral output-to-dose calibration, in vivo scintillator dosimeters exhibited high linearity (R2=0.99) with paired OSLD readings across all patients. The difference between scintillation and OSLD dose measurements averaged 2.8% of the prescribed dose, or an absolute dose difference of approximately 7 cGy. ConclusionsIntegration of scintillation dosimetry with Cherenkov imaging offers an accurate, rapid alternative for in vivo dose verification in WBRT, circumventing the limitations of conventional point dosimeters. The additional benefit of visualizing measurement locations relative to the treatment field provides users an enhanced understanding of results and allows for detection of high dose gradients. Future work will explore the applicability of this technique across a broader range of radiotherapy treatments, aiming to streamline IVD practices.

2453: Dose and setup verification based on in vivo dosimetry for PBS proton RBE mice studies

2453: Dose and setup verification based on in vivo dosimetry for PBS proton RBE mice studies

Review on high spatial resolution dosimetry with pixelated semiconductor detectors for radiation therapy

Current and emerging radiation therapy treatment techniques are characterized by complex and spatially non-uniform dose distributions with steep dose gradients delivered through high-precision dose placement with small radiation fields. Such approaches necessitate advanced verification tools. For accurate dose profile representation in stereotactic radiation therapy using megavoltage X-rays, the Nyquist theorem places the minimum sampling step as 2.5 mm that determines the required spatial resolution of the detectors for dose verification. For techniques such as micro- and mini-beam radiotherapy, micron-scale sampling is required, and sub-millimeter for particle therapy. Silicon semiconductor detectors provide the possibility of producing small sensitive volumes (including in arrays) without compromising mechanical robustness or sensitivity, therefore allowing for high spatial resolution dosimetry. They exhibit many properties of an ideal detector, while the properties of silicon detectors known to negatively influence response can be compensated for and minimized. As such, this work reviews the current status of semiconductor radiation detectors providing dose sampling by diodes with submillimeter sensitive volumes and a pitch of 2.5 mm or smaller for relative dosimetry in X-ray and electron external beam radiation therapy, brachytherapy, microbeam radiation therapy, proton, and heavy ion therapy. The following types of devices are discussed in this work, including commercial systems and those under development: one-dimensional and two-dimensional pixelated arrays (monolithic and single diode based).

Population Pharmacokinetics (PopPK) Support for Pediatric Dosing of Biological Products.

This study assesses the use of population pharmacokinetics (PopPK) in supporting pediatric dosing of novel biological drug products. The labeling for biologic drug products approved by the US Food and Drug Administration (FDA) from 2002 until 2021 was reviewed to identify those with a pediatric indication. For the drugs with a pediatric indication, the dosing regimen(s) based on age groups, dosing strategy, the use of PopPK to support the dose, and the types of pediatric clinical trials were reviewed. Data were collected from FDA's review documents and product labels on the Drugs@FDA website, and as needed, more clinical trial details were collected from PubMed and clinicaltrials.gov. The role of PopPK analyses in dosing was captured when mentioned in the label or review as playing a role in selecting the approved pediatric dose and/or in verifying the adequacy of the studied dose to support labeling. Between 2002 and 2021, FDA approved 169 biological products, and 78 of 169 (46%) products have an approved indication for which the label contains dosing recommendations for pediatric use. For the 78 products approved in pediatrics, there was a total of 180 clinical trials that included pediatric patients. Phase 3 pediatric trials commonly supported pediatric approval and dosing for the reviewed products (64%, 50/78 products; 56.1%, 101/180 trials). PopPK analyses were reported to play a critical role in dose selection, prediction, and verification for 40 of the 78 products (51%), including informing pediatric dosing in the absence of pediatric data (e.g., drugs approved under animal rule), comparing exposures to the exposure range observed in adults, and informing alternative dosing strategies in certain age or body weight groups. PopPK analyses have been applied in a variety of ways to inform pediatric dosing and support extrapolation from adult data or other pediatric age groups for biologics. Understanding and learning from these past cases on the use of pharmacometrics tools to support pediatric dosing of biological products can inform future pediatric development programs.

A feasibility study to predict 3D dose delivery accuracy for IMRT using DenseNet with log files.

This study aims to explore the feasibility of DenseNet in the establishment of a three-dimensional (3D) gamma prediction model of IMRT based on the actual parameters recorded in the log files during delivery. A total of 55 IMRT plans (including 367 fields) were randomly selected. The gamma analysis was performed using gamma criteria of 3% /3 mm (Dose Difference/Distance to Agreement), 3% /2 mm, 2% /3 mm, and 2% /2 mm with a 10% dose threshold. In addition, the log files that recorded the gantry angle, monitor units (MU), multi-leaf collimator (MLC), and jaws position during delivery were collected. These log files were then converted to MU-weighted fluence maps as the input of DenseNet, gamma passing rates (GPRs) under four different gamma criteria as the output, and mean square errors (MSEs) as the loss function of this model. Under different gamma criteria, the accuracy of a 3D GPR prediction model decreased with the implementation of stricter gamma criteria. In the test set, the mean absolute error (MAE) of the prediction model under the gamma criteria of 3% /3 mm, 2% /3 mm, 3% /2 mm, and 2% /2 mm was 1.41, 1.44, 3.29, and 3.54, respectively; the root mean square error (RMSE) was 1.91, 1.85, 4.27, and 4.40, respectively; the Sr was 0.487, 0.554, 0.573, and 0.506, respectively. There was a correlation between predicted and measured GPRs (P < 0.01). Additionally, there was no significant difference in the accuracy between the validation set and the test set. The accuracy in the high GPR group was high, and the MAE in the high GPR group was smaller than that in the low GPR group under four different gamma criteria. In this study, a 3D GPR prediction model of patient-specific QA using DenseNet was established based on log files. As an auxiliary tool for 3D dose verification in IMRT, this model is expected to improve the accuracy and efficiency of dose validation.

Technical note: MR image-based synthesis CT for CyberKnife robotic stereotactic radiosurgery

The purpose of this study is to investigate whether deep learning-based sCT images enable accurate dose calculation in CK robotic stereotactic radiosurgery. A U-net convolutional neural network was trained using 2446 MR-CT pairs and used it to translate 551 MR images to sCT images for testing. The sCT of CK patient was encapsulated into a quality assurance (QA) validation phantom for dose verification. The CT value difference between CT and sCT was evaluated using mean absolute error (MAE) and the statistical significance of dose differences between CT and sCT was tested using the Wilcoxon signed rank test. For all CK patients, the MAE value of the whole brain region did not exceed 25 HU. The percentage dose difference between CT and sCT was less than ±0.4% on GTV (D2(Gy), −0.29%, D95(Gy), −0.09%), PTV (D2(Gy), −0.25%, D95(Gy), −0.10%), and brainstem (max dose(Gy), 0.31%). The percentage dose difference between CT and sCT for most regions of interest (ROIs) was no more than ±0.04%. This study extended MR-based sCT prediction to CK robotic stereotactic radiosurgery, expanding the application scenarios of MR-only radiation therapy. The results demonstrated the remarkable accuracy of dose calculation on sCT for patients treated with CK robotic stereotactic radiosurgery.

Validation of dual-energy CT-based composition analysis using fresh animal tissues and composition-optimized tissue equivalent samples

Objective. Proton therapy allows for highly conformal dose deposition, but is sensitive to range uncertainties. Several approaches currently under development measure composition-dependent secondary radiation to monitor the delivered proton range in-vivo. To fully utilize these methods, an estimate of the elemental composition of the patient’s tissue is often needed. Approach. A published dual-energy computed tomography (DECT)-based composition-extraction algorithm was validated against reference compositions obtained with two independent methods. For this purpose, a set of phantoms containing either fresh porcine tissue or tissue-mimicking samples with known, realistic compositions were imaged with a CT scanner at two different energies. Then, the prompt gamma-ray (PG) signal during proton irradiation was measured with a PG detector prototype. The PG workflow used pre-calculated Monte Carlo simulations to obtain an optimized estimate of the sample’s carbon and oxygen contents. The compositions were also assessed with chemical combustion analysis (CCA), and the stopping-power ratio (SPR) was measured with a multi-layer ionization chamber. The DECT images were used to calculate SPR-, density- and elemental composition maps, and to assign voxel-wise compositions from a selection of human tissues. For a more comprehensive set of reference compositions, the original selection was extended by 135 additional tissues, corresponding to spongiosa, high-density bones and low-density tissues. Results. The root-mean-square error for the soft tissue carbon and oxygen content was 8.5 wt% and 9.5 wt% relative to the CCA result and 2.1 wt% and 10.3 wt% relative to the PG result. The phosphorous and calcium content were predicted within 0.4 wt% and 1.1 wt% of the CCA results, respectively. The largest discrepancies were encountered in samples whose composition deviated the most from tabulated compositions or that were more inhomogeneous. Significance. Overall, DECT-based composition estimations of relevant elements were in equal or better agreement with the ground truth than the established SECT-approach and could contribute to in-vivo dose verification measurements.

Turning the attention to time-resolved EPID-images: treatment error classification with transformer multiple instance learning

Objective. The aim of this work was to develop a novel artificial intelligence-assisted in vivo dosimetry method using time-resolved (TR) dose verification data to improve quality of external beam radiotherapy. Approach. Although threshold classification methods are commonly used in error classification, they may lead to missing errors due to the loss of information resulting from the compression of multi-dimensional electronic portal imaging device (EPID) data into one or a few numbers. Recent research has investigated the classification of errors on time-integrated (TI) in vivo EPID images, with convolutional neural networks showing promise. However, it has been observed previously that TI approaches may cancel out the error presence on γ-maps during dynamic treatments. To address this limitation, simulated TR γ-maps for each volumetric modulated arc radiotherapy angle were used to detect treatment errors caused by complex patient geometries and beam arrangements. Typically, such images can be interpreted as a set of segments where only set class labels are provided. Inspired by recent weakly supervised approaches on histopathology images, we implemented a transformer based multiple instance learning approach and utilized transfer learning from TI to TR γ-maps. Main results. The proposed algorithm performed well on classification of error type and error magnitude. The accuracy in the test set was up to 0.94 and 0.81 for 11 (error type) and 22 (error magnitude) classes of treatment errors, respectively. Significance. TR dose distributions can enhance treatment delivery decision-making, however manual data analysis is nearly impossible due to the complexity and quantity of this data. Our proposed model efficiently handles data complexity, substantially improving treatment error classification compared to models that leverage TI data.

Clinical applicability of Linac-integrated CBCT based NIPAM 3D dosimetry: a dual-institutional investigation

Objective. To develop and benchmark a novel 3D dose verification technique consisting of polymer gel dosimetry (PGD) with cone-beam-CT (CBCT) readout through a two-institution study. The technique has potential for wide and robust applicability through reliance on CBCT readout. Approach. Three treatment plans (3-field, TG119-C-shape spine, 4-target SRS) were created by two independent institutions (Institutions A and B). A Varian Truebeam linear accelerator was used to deliver the plans to NIPAM polymer gel dosimeters produced at both institutions using an identical approach. For readout, a slow CBCT scan mode was used to acquire pre- and post-irradiation images of the gel (1 mm slice thickness). Independent gel analysis tools were used to process the PGD images (A: VistaAce software, B: in-house MATLAB code). Comparing planned and measured doses, the analysis involved a combination of 1D line profiles, 2D contour plots, and 3D global gamma maps (criteria ranging between 2%1 mm and 5%2 mm, with a 10% dose threshold). Main results. For all gamma criteria tested, the 3D gamma pass rates were all above 90% for 3-field and 88% for the SRS plan. For the C-shape spine plan, we benchmarked our 2% 2 mm result against previously published work using film analysis (93.4%). For 2%2 mm, 99.4% (Institution A data), and 89.7% (Institution B data) were obtained based on VistaAce software analysis, 83.7% (Institution A data), and 82.9% (Institution B data) based on MATLAB. Significance. The benchmark data demonstrate that when two institutions follow the same rigorous procedures gamma passing rates up to 99%, for 2%2 mm criteria can be achieved for substantively different treatment plans. The use of different software and calibration techniques may have contributed to the variation in the 3D gamma results. By sharing the data across institutions, we observe the gamma passing rate is more consistent within each pipeline, indicating the need for standardized analysis methods.

Patient-specific deep learning for 3D protoacoustic image reconstruction and dose verification in proton therapy.

Protoacoustic (PA) imaging has the potential to provide real-time 3D dose verification of proton therapy. However, PA images are susceptible to severe distortion due to limited angle acquisition. Our previous studies showed the potential of using deep learning to enhance PA images. As the model was trained using a limited number of patients' data, its efficacy was limited when applied to individualpatients. In this study, we developed a patient-specific deep learning method for protoacoustic imaging to improve the reconstruction quality of protoacoustic imaging and the accuracy of dose verification for individualpatients. Our method consists of two stages: in the first stage, a group model is trained from a diverse training set containing all patients, where a novel deep learning network is employed to directly reconstruct the initial pressure maps from the radiofrequency (RF) signals; in the second stage, we apply transfer learning on the pre-trained group model using patient-specific dataset derived from a novel data augmentation method to tune it into a patient-specific model. Raw PA signals were simulated based on computed tomography (CT) images and the pressure map derived from the planned dose. The reconstructed PA images were evaluated against the ground truth by using the root mean squared errors (RMSE), structural similarity index measure (SSIM) and gamma index on 10 specific prostate cancer patients. The significance level was evaluated by t-test with the p-value threshold of 0.05 compared with the results from the groupmodel. The patient-specific model achieved an average RMSE of 0.014 ( ), and an average SSIM of 0.981 ( ), out-performing the group model. Qualitative results also demonstrated that our patient-specific approach acquired better imaging quality with more details reconstructed when comparing with the group model. Dose verification achieved an average RMSE of 0.011 ( ), and an average SSIM of 0.995 ( ). Gamma index evaluation demonstrated a high agreement (97.4% [ ] and 97.9% [ ] for 1%/3 and 1%/5mm) between the predicted and the ground truth dose maps. Our approach approximately took 6 s to reconstruct PA images for each patient, demonstrating its feasibility for online 3D dose verification for prostate protontherapy. Our method demonstrated the feasibility of achieving 3D high-precision PA-based dose verification using patient-specific deep-learning approaches, which can potentially be used to guide the treatment to mitigate the impact of range uncertainty and improve the precision. Further studies are needed to validate the clinical impact of thetechnique.