Automatic Dose Research Articles

Verification of daily dose recalculation accuracy for an adaptive radiation therapy monitoring tool in helical tomotherapy for nasopharyngeal carcinoma.

PreciseART, an adaptive radiation therapy (ART) software for helical tomotherapy (HT), was integrated into the Precision treatment planning system (Accuray, Inc., Sunnyvale, CA). It supports automatic monitoring of dose variations to both the target and organs at risk (OARs) throughout the treatment course. This study aims to verify the accuracy of PreciseART's automatic dose recalculation and assess the effectiveness of its notification function. The Radixact X9's kVCT image-guided system (ClearRT) was used to acquire daily images for dose recalculations. We assessed the accuracy of PreciseART's automatic dose recalculation by comparing it with the treatment planning system (TPS) recalculation for the PTV70, spinal cord, and bilateral parotid glands. We also evaluated the efficacy of its notification function by comparing each dose metric for each notification color to TPS recalculation, assessing its role as a trigger tool. In the phantom study, dosimetric analysis indicated no statistically significant differences between TPS and PreciseART recalculations (p>0.05), with dose variations under 1.5%. Similarly, in the patient study (n=10), no significant dosimetric discrepancies were found (p>0.05), with a maximum variation of 2.3%. The notification system performed effectively, providing accurate notifications in accordance with predefined dose criteria. PreciseART's daily dose recalculation demonstrated good agreement with TPS recalculation, and its notification function is effective for identifying dose threshold compliance, supporting its use in clinical ART.

Optimizing normal tissue objectives (NTO) in eclipse treatment planning system (TPS) for stereotactic treatment of multiple brain metastases using non-coplanar RapidArc and comparison with HyperArc techniques.

To optimize NTO parameters in non-coplanar RapidArc (RA) stereotactic radiosurgery (SRS) for multiple brain metastases and compare them with HyperArc (HA) plans. Thirty patients with multiple brain metastases, receiving 21Gy prescriptions, were retrospectively enrolled, with lesions ranging from two to eight and volumes between 0.27 and 10.56 cm3. Non-coplanar RapidArc plans utilized manual NTO (RA-mNTO) with varying dose fall-off values (0.1mm-1, 0.5mm-1, 1.0mm-1, 2.0mm-1, 3.0mm-1) and end doses (50%, 25% & 10%). Additionally, two HyperArc plans were generated: HA-ALDO used Automatic Lower Dose Objectives with SRS NTO, while HA-mNTO used the same beam geometry with manual NTO parameters optimized from RA-mNTO plans. TrueBeam with High-Definition Multi-leaf Collimators (HDMLC), 6 MV Flattening Filter Free (FFF) Beam at a maximum dose rate of 1400 MU/min, and Eclipse version 16.1 TPS were used. Plans were assessed for Paddick Conformity Index (CI), Gradient Index (GI), Homogeneity Index (HI), normal brain doses (V18Gy, V15Gy, and V12Gy), Monitor Units (MUs), and delivery accuracy using aS1200 Digital Megavolt Imager (DMI) with 2%/2mm gamma criteria. Statistical analysis utilized integrated scoring and the Wilcoxon signed-rank test. RA-mNTO plans with 0.5mm⁻1 dose fall-off and 10% end-dose showed superior dosimetric outcomes: CI (0.85 ± 0.08), GI (3.63 ± 0.87), and HI (0.36 ± 0.06) compared to HA-ALDO (CI 0.84 ± 0.09, GI: 3.97 ± 0.85, HI: 0.39 ± 0.07) and HA-mNTO (CI 0.83 ± 0.08, GI: 3.60 ± 0.93, HI: 0.40 ± 0.06). MUs were comparable: RA-mNTO (9679 ± 1882), HA-ALDO (9509 ± 1315), and higher for HA-mNTO (10,457 ± 1980). RA-mNTO plans exhibited significantly lower normal brain doses (V18Gy: 1.78 ± 1.23, V15Gy: 3.54 ± 2.37, V12Gy: 6.21 ± 4.09) compared to HA-ALDO (V18Gy: 2.02 ± 1.34, V15Gy: 4.09 ± 2.66, V12Gy: 7.15 ± 4.56) and HA-mNTO (V18Gy: 1.85 ± 1.20, V15Gy: 3.68 ± 2.33, V12Gy: 6.36 ± 3.97). All techniques achieved > 98% gamma pass rate. Non-coplanar RA plans with optimized mNTO settings outperformed HyperArc plans in all studied dosimetric parameters for SRS treatment of multiple brain metastases.

A novel method for accurate and real-time quantification of radiopharmaceutical activity in automated dose dispensers

In recent years, there has been a rapid growth in new technologies within nuclear medicine and PET diagnostics, particularly in the field of radiopharmaceutical dose preparation. These advancements are driven by the integration of automatic algorithms into the process. This evolution has brought about significant improvements in the efficiency and accuracy of radiopharmaceutical dose preparation. These automatic algorithms utilize state-of-the-art technologies to streamline the process and ensure precise and tailored doses for patients. The integration of these advanced technologies has revolutionized the field, allowing for more effective and personalized treatment options in nuclear medicine and PET diagnostics. By employing automatic radiopharmaceutical dose dispensers (ARDDs), radiation exposure to technologists is minimized while also mitigating the risk of potential infections for patients. Ionization chambers are currently the most used detectors to measure the activity of radiopharmaceuticals in ARDDs. However, they present certain limitations, including size constraints. This study is part of an international endeavor to develop a new approach for evaluating radiopharmaceutical activity in an ARDD named KARl 100. A new approach was used to develop a scintillation detector by linking a PVT detector to a SiPMs array consisting of four elements, each having an active surface of 6 × 6 mm with a bias voltage of 30 V. The detection system's amplified output was then fed to a single-channel analyzer called Enviro, developed by Tema Sinergie S.p.A in Italy. The counts per second (CPS) detected by the detector were studied using WinTAM software allowing for the radionuclide’s activity concentration to be read out at run-time. Three widely used medical radioisotopes, 18F, 11C and 68Ga, were used to evaluate the performance of the detector in terms of linearity and repeatability of the measurements. Half-life and linear response were evaluated over a wide range of activities with different decay schemes. Data analysis was conducted, resulting in half-lives of the 18F, 11C and 68Ga radioisotopes of 109.75 ± 0.35, 20.62 ± 0.34 and 67.82 ± 0.25 min, respectively, with calculated deviations of 0.0167%, 1.00% and 0.29% from expected values. The linear response of the detector over the entire range of desired activity concentrations was confirmed after data correction for the detector dead time. Calibration of the detector combined with the dispenser system was carried out in parallel investigations to provide the required device properties for reference applications in nuclear medicine. Furthermore, Monte Carlo simulations using FLUKA code were performed to investigate the contribution of beta particles in addition to gamma radiations to assess the influence of beta particles on the calibration factors. For better accuracy, a shielding was designed to prevent any beta particles from reaching the detector.

Beam-wise dose composition learning for head and neck cancer dose prediction in radiotherapy

Automatic and accurate dose distribution prediction plays an important role in radiotherapy plan. Although previous methods can provide promising performance, most methods did not consider beam-shaped radiation of treatment delivery in clinical practice. This leads to inaccurate prediction, especially on beam paths. To solve this problem, we propose a beam-wise dose composition learning (BDCL) method for dose prediction in the context of head and neck (H&N) radiotherapy plan. Specifically, a global dose network is first utilized to predict coarse dose values in the whole-image space. Then, we propose to generate individual beam masks to decompose the coarse dose distribution into multiple field doses, called beam voters, which are further refined by a subsequent beam dose network and reassembled to form the final dose distribution. In particular, we design an overlap consistency module to keep the similarity of high-level features in overlapping regions between different beam voters. To make the predicted dose distribution more consistent with the real radiotherapy plan, we also propose a dose-volume histogram (DVH) calibration process to facilitate feature learning in some clinically concerned regions. We further apply an edge enhancement procedure to enhance the learning of the extracted feature from the dose falloff regions. Experimental results on a public H&N cancer dataset from the AAPM OpenKBP challenge show that our method achieves superior performance over other state-of-the-art approaches by significant margins. Source code is released at https://github.com/TL9792/BDCLDosePrediction.

Optimization of Radiation Protection in Pediatric Interventional Radiology in Latin America and the Caribbean: Development, Advancements, Challenges and Achievements of the OPRIPALC Program.

This article presents the development, advancements, challenges and achievements of the "Optimization of Protection in Pediatric Interventional Radiology in Latin America and the Caribbean" (OPRIPALC) program. This international initiative is led by the World Health Organization, the Pan American Health Organization and the International Atomic Energy Agency. The main objectives of OPRIPALC are to foster a culture of radiological protection in pediatric interventions, enhance these procedures' quality, and define optimization strategies such as the use of diagnostic reference levels (DRLs). Currently, 33 centers from 12 countries participate actively in the program. Significant progress has been made towards the proposed objectives, overcoming the challenges posed by the COVID-19 pandemic. Through many virtual meetings for coordination, planning, training and follow-up, a comprehensive set of DRLs for both diagnostic and therapeutic procedures, categorized by weight and age, have been established and are in use. A consensus document on good practices is in the final stage of development. The program's continuation into at least a second phase is essential to address pending issues, including the integration of automatic dose management systems, the levels of occupational radiation doses, their correlation with pediatric patient doses, and strategies to reduce them.

Comparison of Ethos template-based planning and AI-based dose prediction: General performance, patient optimality, and limitations

PurposeEthos proposes a template-based automatic dose planning (Etb) for online adaptive radiotherapy. This study evaluates the general performance of Etb for prostate cancer, as well as the ability to generate patient-optimal plans, by comparing it with another state-of-the-art automatic planning method, i.e., deep learning dose prediction followed by dose mimicking (DP + DM). MaterialsGeneral performances and capability to produce patient-optimal plan were investigated through two studies: Study-S1 generated plans for 45 patients using our initial Ethos clinical goals template (EG_init), and compared them to manually generated plans (MG). For study-S2, 10 patients which showed poor performances at study-S1 were selected.S2 compared the quality of plans generated with four different methods: 1) Ethos initial template (EG_init_selected), 2) Ethos updated template-based on S1 results (EG_upd_selected), 3) DP + DM, and 4) MG plans. ResultsEG_init plans showed satisfactory performance for dose level above 50 Gy: reported mean metrics differences (EG_init minus MG) never exceeded 0.6 %. However, lower dose levels showed loosely optimized metrics, mean differences for V30Gy to rectum and V20Gy to anal canal were of 6.6 % and 13.0 %. EG_init_selected showed amplified differences in V30Gy to rectum and V20Gy to anal canal: 8.5 % and 16.9 %, respectively. These dropped to 5.7 % and 11.5 % for EG_upd_selected plans but strongly increased V60Gy to rectum for 2 patients. DP + DM plans achieved differences of 3.4 % and 4.6 % without compromising any V60Gy. ConclusionGeneral performances of Etb were satisfactory. However, optimizing with template of goals might be limiting for some complex cases. Over our test patients, DP + DM outperformed the Etb approach.

Hyperarc Vmat and Vmat planning for stereotactic radiosurgery in multiple brain metastases

Automated treatment planning systems are available for linear accelerator (LINAC) based single and more isocenter multitarget stereotactic radiosurgery (SRS) of brain metastases. HyperArc Volumetric Modulated Arc Therapy (HA-VMAT) is a relatively new isocentric Volumetric Modulated Arc Therapy (VMAT) technique developed specifically for non-coplanar, multileaf collimator (MLC)-based stereotactic radiotherapy with automatic treatment optimization and dose delivery. The HA-VMAT planning approach was developed to accommodate dosing procedures for brain metastases stereotactic radiosurgery. In this study, it was aimed to compare the dosimetric parameters of treatment plans created with single and multi-isocenter retrospectively VMAT and HA-VMAT techniques in patients with lung cancer and multiple brain metastases (4–7) who received single and multi-fraction stereotactic radiosurgery (SRS). Using VMAT and HA-VMAT techniques, two separate treatment plans were created for each case, and tumor target coverage (homogeneity index (HI); conformity index (CI); gradient index (GI)) and organ at risk (OAR) were evaluated dosimetrically. Physical properties and Monitor unit (MU) values of both treatment approaches were compared. Treatment plans were created for study purposes and patients were not treated with these plans. In the study, HI (p = 0.000), CI (p = 0.000) and GI (p = 0.002) values calculated with the HA-VMAT technique are superior to the VMAT technique. Medium-low dose spreads (V4Gy-V16Gy) were significantly reduced in HA-VMAT plans compared to plans with VMAT. HA-VMAT plans result in lower MU than VMAT plans. HA-VMAT planning has significantly less radiation necrosis indicator (V8Gy-V16Gy) compared to VMAT plans. However, there are few studies investigating the ability of the HA-VMAT technique to create high-quality treatment plans for extracranial lesions. HA-VMAT may constitute one of the new options for SRS dose delivery for both single target and multiple targets.

Computed tomography dose index determination in dose modulation prospectively involving the third-generation iterative reconstruction and noise index.

Optimizing CT protocols is challenging in the presence of automatic dose modulation because the CT dose index (CTDIvol) at different patient sizes is unknown to the operator. The task is more difficult when both the image quality index and iterative reconstruction prospectively affect the dose determination. It is of interest in practice to be informed of the CTDIvol during the protocol initialization and evaluation. It was our objective to obtain a predictive relationship between CTDIvol, the image quality index, and iterative reconstruction strength at various patient sizes. Dose modulation data were collected on a GE Revolution 256-slice scanner utilizing a Mercury phantom and selections of the noise index (NI) from 8 to 17, the third generation iterative reconstruction (ASIR-V) from 0% to 80%, and phantom diameters from 16 to 36cm. The fixed parameters were 120kVp, a pitch of .984, and a collimation of 40mm with a primary slice width of 2.5mm. The CTDIvol per diameter was based on the average tube current over three adjacent slices (same or similar diameter) multiplied by a conversion factor between the average mA of the series and the reported CTDIvol. The relationship between CTDIvol, NI, and ASIR-V for each diameter was fitted with a 2nd order polynomial of ASIR-V multiplied by a power law of NI. The ASIR-V fit parameters versus diameter followed a Lorentz function while the NI exponent versus diameter followed an exponential growth function. The CTDIvol predictions were accurate within 15% compared to phantom results on a separate GE Revolution. For clinical relevance, the phantom diameter was converted to an abdomen or chest equivalent diameter and was well matched to patient data. The fitted relationship for CTDIvol. for given values of NI and ASIR-V blending for a range of phantom sizes was a good match to phantom and patient data. The results can be of direct help for selecting adequate parameters in CT protocol development.

TransDose: Transformer-based radiotherapy dose prediction from CT images guided by super-pixel-level GCN classification.

Radiotherapy is a mainstay treatment for cancer in clinic. An excellent radiotherapy treatment plan is always based on a high-quality dose distribution map which is produced by repeated manual trial-and-errors of experienced experts. To accelerate the radiotherapy planning process, many automatic dose distribution prediction methods have been proposed recently and achieved considerable fruits. Nevertheless, these methods require certain auxiliary inputs besides CT images, such as segmentation masks of the tumor and organs at risk (OARs), which limits their prediction efficiency and application potential. To address this issue, we design a novel approach named as TransDose for dose distribution prediction that treats CT images as the unique input in this paper. Specifically, instead of inputting the segmentation masks to provide the prior anatomical information, we utilize a super-pixel-based graph convolutional network (GCN) to extract category-specific features, thereby compensating the network for the necessary anatomical knowledge. Besides, considering the strong continuous dependency between adjacent CT slices as well as adjacent dose maps, we embed the Transformer into the backbone, and make use of its superior ability of long-range sequence modeling to endow input features with inter-slice continuity message. To our knowledge, this is the first network that specially designed for the task of dose prediction from only CT images without ignoring necessary anatomical structure. Finally, we evaluate our model on two real datasets, and extensive experiments demonstrate the generalizability and advantages of our method.

Unsupervised Domain Adaptive Dose Prediction Via Cross-Attention Transformer and Target-Specific Knowledge Preservation.

Radiotherapy is one of the leading treatments for cancer. To accelerate the implementation of radiotherapy in clinic, various deep learning-based methods have been developed for automatic dose prediction. However, the effectiveness of these methods heavily relies on the availability of a substantial amount of data with labels, i.e. the dose distribution maps, which cost dosimetrists considerable time and effort to acquire. For cancers of low-incidence, such as cervical cancer, it is often a luxury to collect an adequate amount of labeled data to train a well-performing deep learning (DL) model. To mitigate this problem, in this paper, we resort to the unsupervised domain adaptation (UDA) strategy to achieve accurate dose prediction for cervical cancer (target domain) by leveraging the well-labeled high-incidence rectal cancer (source domain). Specifically, we introduce the cross-attention mechanism to learn the domain-invariant features and develop a cross-attention transformer-based encoder to align the two different cancer domains. Meanwhile, to preserve the target-specific knowledge, we employ multiple domain classifiers to enforce the network to extract more discriminative target features. In addition, we employ two independent convolutional neural network (CNN) decoders to compensate for the lack of spatial inductive bias in the pure transformer and generate accurate dose maps for both domains. Furthermore, to enhance the performance, two additional losses, i.e. a knowledge distillation loss (KDL) and a domain classification loss (DCL), are incorporated to transfer the domain-invariant features while preserving domain-specific information. Experimental results on a rectal cancer dataset and a cervical cancer dataset have demonstrated that our method achieves the best quantitative results with [Formula: see text], [Formula: see text], and HI of 1.446, 1.231, and 0.082, respectively, and outperforms other methods in terms of qualitative assessment.

Automatic dose prediction using deep learning and plan optimization with finite-element control for intensity modulated radiation therapy.

Automatic solutions for generating radiotherapy treatment plans using deep learning (DL) have been investigated by mimicking the voxel's dose. However, plan optimization using voxel-dose features has not been extensively studied. This study aims to investigate the efficiency of a direct optimization strategy with finite elements (FEs) after DL dose prediction for automatic intensity-modulated radiation therapy (IMRT) treatment planning. A double-UNet DL model was adapted for 220 cervical cancer patients (200 for training and 20 for testing), who underwent IMRT between 2016 and 2020 at our clinic. The model inputs were computed tomography (CT) slices, organs at risk (OARs), and planning target volumes (PTVs), and the outputs were dose distributions of uniformly generated high-dose region-controlled plans. The FEs were discretized into equal intervals of the dose prediction value within the [OARs avoid PTV(O-P)] and [body avoids OARs & PTV(B-OP)] regions in the test cohort and used to define the objectives for IMRT plan optimization. The plans were optimized using a two-step process. In the beginning, the plans of two extra cases with and without low-dose region control were compared to pursue robust and optimal dose adjustment degree pattern of FEs. In the first step, the mean dose of O-P FEs were constrained to differing degrees according to the pattern. The further the FEs from the PTV, the tighter the constraints. In the second step, the mean dose of O-P FEs from first step were constrained again but weakly and the dose of the B-OP FEs from dose prediction and PTV were tightly regulated. The dosimetric parameters of the OARs and PTV were evaluated and compared using an interstep approach. In another 10 cases, the plans optimized via the aforementioned steps (method 1) were compared with those directly generated by the double-UNet dose prediction model trained by low and high region-controlled plans (method 2). The mean differences in dose metrics between the UNet-predicted dose and the clinical plans were: 0.47Gy for bladder D50% ; 0.62Gy for rectum D50% ; 0% for small intestine V30Gy ; 1% for small intestine V40Gy ; 4% for left femoral head V30Gy ; and 6% for right femoral head V30Gy . The reductions in mean dose (p<0.001) after FE-based optimization were: 4.0, 1.9, 2.8, 5.9, and 5.7Gy for the bladder, rectum, small intestine, left femoral head, and right femoral head, respectively, with flat PTV homogeneity and conformity. Method 1 plans produced lower mean doses than those of method 2 for the bladder (0.7Gy), rectum (1.0Gy), and small intestine (0.6Gy), while maintaining PTV homogeneity and conformity. FE-based direct optimization produced lower OAR doses and adequate PTV doses after DL prediction. This solution offers rapid and automatic plan optimization without manual adjustment, particularly in low-dose regions.

Effect of shape of automatic dose rate control and wedge compensation filter on radiation dose in an angiography system with a flat-panel detector.

The purpose was to investigate air-kerma area product (PKA) and entrance surface air-kerma rate ([Formula: see text]a,e) on the effect of the shape of automatic dose rate control (ADRC) in the presence of a wedge compensation filter. We compared and evaluated the variability of the X-ray output using a combination of wedge compensation filters and the ADRC. Two ADRC shapes (round and square) and three poly-methyl-methacrylate thicknesses (15, 20, and 25cm) were used. A wedge compensation filter was inserted 2cm at a time, up to 6cm. When the wedge compensation filter was inserted to 6 cm for 20cm of poly-methyl-methacrylate, the X-ray output fluctuated significantly. The PKA was reduced by 39% when the wedge compensation filter was inserted to 6 cm and by 59% when it was inserted to 4cm under round-type for 20cm poly-methyl-methacrylate. The shape of the ADRC affects [Formula: see text]a,e and PKA.

Coronary Computed Tomography Angiography with Deep Learning Image Reconstruction: A Preliminary Study to Evaluate Radiation Exposure Reduction.

Coronary computed tomography angiography (CCTA) is a medical imaging technique that produces detailed images of the coronary arteries. Our work focuses on the optimization of the prospectively ECG-triggered scan technique, which delivers the radiation efficiently only during a fraction of the R-R interval, matching the aim of reducing radiation dose in this increasingly used radiological examination. In this work, we analyzed how the median DLP (Dose-Length Product) values for CCTA of our Center decreased significantly in recent times mainly due to a notable change in the technology used. We passed from a median DLP value of 1158 mGy·cm to 221 mGy·cm for the whole exam and from a value of 1140 mGy·cm to 204 mGy·cm if considering CCTA scanning only. The result was obtained through the association of important factors during the dose imaging optimization: technological improvement, acquisition technique, and image reconstruction algorithm intervention. The combination of these three factors allows us to perform a faster and more accurate prospective CCTA with a lower radiation dose. Our future aim is to tune the image quality through a detectability-based study, combining algorithm strength with automatic dose settings.

AN INSIGHT DATA ON MEDICATION SAFETY IN ACCORDANCE WITH DRUG COMPLICATIONS BASED ON THE AUTOMATED DISPENSING SYSTEM

Automated dispensing systems have been applied for maintaining medication safety, improving drug distribution, and reducing the risk of medication errors in different pharmacy practice. Automated dispensing machines maintain the balance of accessibility, stock control of medications, and protection which were the main features of a safe medication distribution system. It should be linked with medication review and regular patient counselling. Encourage the timely administration of drugs through expanding their access to patient care units and first-dose availability was improved in automated dispensing machines. Physicians allow treatment efficiently by providing convenient access to medications of critically ill patients for emergencies during and after pharmacy hours with automated dispensing machines. This investigation demonstrates that the nature of pharmacotherapy for patients with automated dose dispensing can be improved. This study highlights the impact of an automated dispensing system in primary healthcare, Automated Dispensing Cabinets, or, in an intensive care unit to reduce medication errors. Implementation of automated dispensing systems reduces dispensing errors and gives pharmacists more time to review patient profiles. It recommends that every patient with automatic dose dispensing ought to go through a careful medication review through prescribers and pharmacists. It concluded that the automatic drug dispensing focuses on the effect of a medication review in patients. The ideal recurrence for conducting medication reviews and follow-up will likely contrast between individual patients. Hence, it provides an insight data on medication safety based on the automated dispensing system in accordance with drug complications. KEYWORDS: Automated dose dispensing, Medication safety, Drug complication, Impact, Evaluation, Insight report on medication safety

Direct Dose Prediction With Deep Learning for Postoperative Cervical Cancer Underwent Volumetric Modulated Arc Therapy.

To predict the voxel-based dose distribution for postoperative cervical cancer patients underwent volumetric modulated arc therapy using deep learning models. A total of 254 patients with cervical cancer received volumetric modulated arc therapy in authors' hospital from January 2018 to September 2021 were enrolled in this retrospective study. Two deep learning networks (3D deep residual neural network and 3DUnet) were adapted to train (203 cases) and test (51 cases) the feasibility and effectiveness of the prediction method. The performance of deep learning models was evaluated by comparing the results with those of treatment planning system according to metrics of dose-volume histogram of target volumes and organs at risk. The dose distributions predicted by deep learning models were clinically acceptable. The automatic dose prediction time was around 5 to 10 min, which was about one-eighth to one-tenth of the manual optimization time. The maximum dose difference was observed in D98 of rectum with a | δD| of 5.00 ± 3.40% and 4.88 ± 3.99% for Unet3D and ResUnet3D, respectively. The minimum difference was observed in the D2 of clinical target volume with a |δD| of 0.53 ± 0.45% and 0.83 ± 0.45% for ResUnet3D and Unet3D, respectively. The 2 deep learning models adapted in the study showed the feasibility and reasonable accuracy in the voxel-based dose prediction for postoperative cervical cancer underwent volumetric modulated arc therapy. Automatic dose distribution prediction of volumetric modulated arc therapy with deep learning models is of clinical significance for the postoperative management of patients with cervical cancer.

Prediction of Radiation Therapy Dose for Lung Cancer IMRT Technique using Support Vector Regression Model

Optimal dose distribution in the treatment planning system (TPS) is crucial before being applied to radiotherapy patients. However, TPS still uses optimization methods that are time-consuming and user-dependent. This study aimed to evaluate the automatic dose prediction model, support vector regression (SVR), and compare it with the clinically planned dose of lung cancer patients. Sixty patients treated with intensity-modulated radiation therapy (IMRT) were used as the objects in this study. The target dose distribution was evaluated based on the conformity index (CI), and dose homogeneity was evaluated with the homogeneity index (HI). In contrast, the mean and maximum doses were used to evaluate organs at risk (right lung, left lung, heart, and spinal cord). Statistical analysis was performed using the Wilcoxon test. A value of <0.05 indicates a significant difference between the two datasets. The mean CI of the SVR and clinical are 1.154±0.003 and 1.181±0.136. The mean HI for SVR and clinical was 0.075±0.016 and 0.083±0.030. the Wilcoxon test showed no statistically significant difference between the two results. The maximum cardiac dose showed a statistically significant difference (p=0.042), while the mean dose and maximum dose of other OARs did not show a statistically significant difference. The study showed no significant difference between the two strategies, except for the maximum heart dose. The model provides information about dose distribution that can be applied clinically to determine the best technique to use in patients.

Deep learning-based low-dose CT for adaptive radiotherapy of abdominal and pelvic tumors

The shape and position of abdominal and pelvic organs change greatly during radiotherapy, so image-guided radiation therapy (IGRT) is urgently needed. The world’s first integrated CT-linac platform, equipped with fan beam CT (FBCT), can provide a diagnostic-quality FBCT for achieve adaptive radiotherapy (ART). However, CT scans will bring the risk of excessive scanning radiation dose. Reducing the tube current of the FBCT system can reduce the scanning dose, but it will lead to serious noise and artifacts in the reconstructed images. In this study, we proposed a deep learning method, Content-Noise Cycle-Consistent Generative Adversarial Network (CNCycle-GAN), to improve the image quality and CT value accuracy of low-dose FBCT images to meet the requirements of adaptive radiotherapy. We selected 76 patients with abdominal and pelvic tumors who received radiation therapy. The patients received one low-dose CT scan and one normal-dose CT scan in IGRT mode during different fractions of radiotherapy. The normal dose CT images (NDCT) and low dose CT images (LDCT) of 70 patients were used for network training, and the remaining 6 patients were used to validate the performance of the network. The quality of low-dose CT images after network restoration (RCT) were evaluated in three aspects: image quality, automatic delineation performance and dose calculation accuracy. Taking NDCT images as a reference, RCT images reduced MAE from 34.34 ± 5.91 to 20.25 ± 4.27, PSNR increased from 34.08 ± 1.49 to 37.23 ± 2.63, and SSIM increased from 0.92 ± 0.08 to 0.94 ± 0.07. The P value is less than 0.01 of the above performance indicators indicated that the difference were statistically significant. The Dice similarity coefficients (DCS) between the automatic delineation results of organs at risk such as bladder, femoral heads, and rectum on RCT and the results of manual delineation by doctors both reached 0.98. In terms of dose calculation accuracy, compared with the automatic planning based on LDCT, the difference in dose distribution between the automatic planning based on RCT and the automatic planning based on NDCT were smaller. Therefore, based on the integrated CT-linac platform, combined with deep learning technology, it provides clinical feasibility for the realization of low-dose FBCT adaptive radiotherapy for abdominal and pelvic tumors.

Automatic dose verification system for breast radiotherapy: Method validation, contour propagation and DVH parameters evaluation

PurposeImage guided radiotherapy (IGRT) strategies allow detecting and monitoring anatomical changes during external beam radiotherapy (EBRT). However, assessing the dosimetric impact of anatomical changes is not straightforward. In current IGRT strategies dose volume histograms (DVH) are not available due to lack of contours and dose recalculations on the cone-beam CT (CBCT) scan. This study investigates the feasibility of using automatically calculated DVH parameters in CBCTs using an independent dose calculation engine and propagated contours. MethodA prospective study (NCT03385031) of thirty-one breast cancer patients who received additional CBCT imaging (N = 70) was performed. Manual and automatically propagated contours were generated for all CBCTs and an automatic dose recalculation was performed. Differences between planned and CBCT-derived DVH parameters (mean and maximum dose to targets, 95% volume coverage to targets and mean heart dose (MHD)) were calculated using the dose verification system with manual and propagated contours and, in both cases, benchmarked against DVH differences quantified in the TPS using manually contoured CBCTs. ResultsDifferences in DVH parameters between the TPS and dose verification system with propagated contours were −1.3% to 0.7% (95% CI) for mean dose to the target volume, −0.3 to 0.2 Gy (95% CI) in MHD and −3.9% to 2.9% (95% CI) in target volume coverage. ConclusionThe use of an independent fully automatic dose verification system with contour propagation showed to be feasible and sufficiently reliable to recalculate CBCT based DVHs during breast EBRT. Volume coverage parameters, i.e. V95%, proved to be especially sensitive to contouring differences.

Evaluation of the clinical value of automatic online dose restoration for adaptive proton therapy of head and neck cancer

IntroductionIntensity modulated proton therapy (IMPT) is highly sensitive to anatomical variations which can cause inadequate target coverage during treatment. This study compares not-adapted (NA) robust plans to two adaptive IMPT methods - a fully-offline adaptive (FOA) and a simplified automatic online adaptive strategy (dose restoration (DR)) to determine the benefit of DR, in head and neck cancer (HNC). Material/methodsRobustly optimized clinical IMPT doses in planning-CTs (pCTs) were available for a cohort of 10 HNC patients. During robust re-optimization, DR used isodose contours, generated from the clinical dose on pCTs, and patient specific objectives to reproduce the clinical dose in every repeated-CT(rCT). For each rCT(n = 50), NA, DR and FOA plans were robustly evaluated. ResultsAn improvement in DVH-metrics and robustness was seen for DR and FOA plans compared to NA plans. For NA plans, 74%(37/50) of rCTs did not fulfill the CTV coverage criteria (D98%>95%Dprescription). DR improved target coverage, target homogeneity and variability on critical risk organs such as the spinal cord. After DR, 52%(26/50) of rCTs met all clinical goals. Because of large anatomical changes and/or inaccurate patient repositioning, 48%(24/50) of rCTs still needed full offline adaptation to ensure an optimal treatment since dose restoration was not able to re-establish the initial plan quality. ConclusionRobust optimization together with fully-automatized DR avoided offline adaptation in 52% of the cases. Implementation of dose restoration in clinical routine could ensure treatment plan optimality while saving valuable human and material resources to radiotherapy departments.

Evaluation of a Generative Adversarial Network to Improve Image Quality and Reduce Radiation-Dose during Digital Breast Tomosynthesis.

In this study, we evaluated the improvement of image quality in digital breast tomosynthesis under low-radiation dose conditions of pre-reconstruction processing using conditional generative adversarial networks [cGAN (pix2pix)]. Pix2pix pre-reconstruction processing with filtered back projection (FBP) was compared with and without multiscale bilateral filtering (MSBF) during pre-reconstruction processing. Noise reduction and preserve contrast rates were compared using full width at half-maximum (FWHM), contrast-to-noise ratio (CNR), peak signal-to-noise ratio (PSNR), and structural similarity (SSIM) in the in-focus plane using a BR3D phantom at various radiation doses [reference-dose (automatic exposure control reference dose: AECrd), 50% and 75% reduction of AECrd] and phantom thicknesses (40 mm, 50 mm, and 60 mm). The overall performance of pix2pix pre-reconstruction processing was effective in terms of FWHM, PSNR, and SSIM. At ~50% radiation-dose reduction, FWHM yielded good results independently of the microcalcification size used in the BR3D phantom, and good noise reduction and preserved contrast. PSNR results showed that pix2pix pre-reconstruction processing represented the minimum in the error with reference FBP images at an approximately 50% reduction in radiation-dose. SSIM analysis indicated that pix2pix pre-reconstruction processing yielded superior similarity when compared with and without MSBF pre-reconstruction processing at ~50% radiation-dose reduction, with features most similar to the reference FBP images. Thus, pix2pix pre-reconstruction processing is promising for reducing noise with preserve contrast and radiation-dose reduction in clinical practice.