Three-dimensional Dose Distributions Research Articles

Enhancing stereotactic ablative boost radiotherapy dose prediction for bulky lung cancer: A multi-scale dilated network approach with scale-balanced structure loss.

Partial stereotactic ablative boost radiotherapy (P-SABR) effectively treats bulky lung cancer; however, the planning process for P-SABR requires repeated dose calculations. To improve planning efficiency, we proposed a novel deep learning method that utilizes limited data to accurately predict the three-dimensional (3D) dose distribution of the P-SABR plan for bulky lung cancer. We utilized data on 74 patients diagnosed with bulky lung cancer who received P-SABR treatment. The patient dataset was randomly divided into a training set (51 plans) with augmentation, validation set (7 plans), and testing set (16 plans). We devised a 3D multi-scale dilated network (MD-Net) and integrated a scale-balanced structure loss into the loss function. A comparative analysis with a classical network and other advanced networks with multi-scale analysis capabilities and other loss functions was conducted based on the dose distributions in terms of the axial view, average dose scores (ADSs), and average absolute differences of dosimetric indices (AADDIs). Finally, we analyzed the predicted dosimetric indices against the ground-truth values and compared the predicted dose-volume histogram (DVH) with the ground-truth DVH. Our proposed dose prediction method for P-SABR plans for bulky lung cancer demonstrated strong performance, exhibiting a significant improvement in predicting multiple indicators of regions of interest (ROIs), particularly the gross target volume (GTV). Our network demonstrated increased accuracy in most dosimetric indices and dose scores in different ROIs. The proposed loss function significantly enhanced the predictive performance of the dosimetric indices. The predicted dosimetric indices and DVHs were equivalent to the ground-truth values. Our study presents an effective model based on limited datasets, and it exhibits high accuracy in the dose prediction of P-SABR plans for bulky lung cancer. This method has potential as an automated tool for P-SABR planning and can help optimize treatments and improve planning efficiency.

Nested CNN architecture for three-dimensional dose distribution prediction in tomotherapy for prostate cancer.

The hypothesis of changing network layers to increase the accuracy of dose distribution prediction, instead of expanding their dimensions, which requires complex calculations, has been considered in our study. Atotal of 137 prostate cancer patients treated with the tomotherapy technique were categorized as 80% training and validating as well as 20% testing for the nested UNet and UNet architectures. Mean absolute error (MAE) was used to measure the dosimetry indices of dose-volume histograms (DVHs), and geometry indices, including the structural similarity index measure (SSIM), dice similarity coefficient (DSC), and Jaccard similarity coefficient (JSC), were used to evaluate the isodose volume (IV) similarity prediction. To verify a statistically significant difference, the two-way statistical Wilcoxon test was used at alevel of 0.05 (p < 0.05). Use of anested UNet architecture reduced the predicted dose MAE in DVH indices. The MAE for planning target volume (PTV), bladder, rectum, and right and left femur were D98% = 1.11 ± 0.90; D98% = 2.27 ± 2.85, Dmean = 0.84 ± 0.62; D98% = 1.47 ± 12.02, Dmean = 0.77 ± 1.59; D2% = 0.65 ± 0.70, Dmean = 0.96 ± 2.82; and D2% = 1.18 ± 6.65, Dmean = 0.44 ± 1.13, respectively. Additionally, the greatest geometric similarity was observed in the mean SSIM for UNet and nested UNet (0.91 vs.0.94, respectively). The nested UNet network can be considered asuitable network due to its ability to improve the accuracy of dose distribution prediction compared to the UNet network in an acceptable time.

Visualization of spatial dose distribution for effective radiation protection education in interventional radiology: obtaining high-accuracy spatial doses.

In recent years, eye lens exposure among radiation workers has become a serious concern in medical X-ray fluoroscopy and interventional radiology (IVR), highlighting the need for radiation protection education and training. This study presents a method that can maintain high accuracy when calculating spatial dose distributions obtained via Monte Carlo simulation and establishes another method to three-dimensionally visualize radiation using the obtained calculation results for contributing to effective radiation-protection education in X-ray fluoroscopy and IVR. The Monte Carlo particle and heavy ion transport code system (PHITS, Ver. 3.24) was used for calculating the spatial dose distribution generated by an angiography device. We determined the peak X-ray tube voltage and half value layer using Raysafe X2 to define the X-ray spectrum from the source and calculated the X-ray spectrum from the measured results using an approximation formula developed by Tucker et al. Further, we performed measurements using the "jungle-gym" method under the same conditions as the Monte Carlo calculations for verifying the accuracy of the latter. An optically stimulated luminescence dosimeter (nanoDot dosimeter) was used as the measuring instrument. In addition, we attempted to visualize radiation using ParaView (version 5.12.0-RC2) using the spatial dose distribution confirmed by the above calculations. A comparison of the measured and Monte Carlo calculated spatial dose distributions revealed that some areas showed large errors (12.3 and 24.2%) between the two values. These errors could be attributed to the scattering and absorption of X-rays caused by the jungle gym method, which led to uncertain measurements, and (2) the angular and energy dependencies of the nanoDot dosimetry. These two causes explain the errors in the actual values, and thus, the Monte Carlo calculations proposed in this study can be considered to have high-quality X-ray spectra and high accuracy. We successfully visualized the three-dimensional spatial dose distribution for direct and scattered X-rays separately using the obtained spatial dose distribution. We established a method to verify the accuracy of Monte Carlo calculations performed through the procedures considered in this study. Various three-dimensional spatial dose distributions were obtained with assured accuracy by applying the Monte Carlo calculation (e.g., changing the irradiation angle and adding a protective plate). Effective radiation-protection education can be realized by combining the present method with highly reliable software to visualize dose distributions.

3D printed dosimeter incorporating leuco-crystal violet and PMMA

This paper describes the development of 3D-printed dosimeters made from radiochromic materials. Incorporating polymethyl methacrylate (PMMA) as a transparent polymer matrix and leuco-crystal violet as a radiochromic dye, these dosimeters exhibit a color change when exposed to radiation. This change provides a way to visualize and measure three-dimensional dose distributions, which is important for the accurate verification of radiation doses in radiotherapy. The use of 3D printing technology allows these dosimeters to approximate human organ geometries, which may contribute to safer and more effective radiotherapeutic applications. The dosimeters can be customized into various shapes and sizes, are lightweight, and cost-effective, making them appropriate for use in both clinical and research settings in radiotherapy. The fabrication process is detailed, and the dosimeters have been tested at doses up to 100 Gy in X-ray irradiation and analyzed for dose distribution. Bragg-peak measurements from carbon beam irradiation illustrate the dosimeters' capability to detect peak radiation doses with a spatial resolution of approximately 1 mm. These findings indicate the potential of these 3D printed dosimeters to improve the accuracy of radiation delivery, which could positively impact patient outcomes in radiotherapy.

An integrated system for measuring 3D dose distributions of carbon-ion pencil beams in regular QA practice

The accurate measurement of three-dimensional (3D) dose distribution of carbon-ion pencil beams in water is crucial for regular quality assurance (QA) practice. Although ionization chambers scanned in a water tank is conventionally used for this purpose, these measuring methods are time-consuming and labor-intensive. In addition, the beam of Heavy Ion Medical Machine (HIMM) changes continuously in each extracted duration, these methods cannot obtain the accurate 3D dose distribution. To solve these problems, an integrated system (PBDOSE) is conceived and established, which employs multi-strip ionization chambers (MSICs) to measure 2D lateral profile dose distributions. The lateral profile dose distribution in the context of various depths is measured by a mobile MSIC and normalized by the measurement of a reference MSIC to reduce the effect of beam instability. On this basis, a 3D dose distribution is created by stacking those profiles in depth direction. By conducting a single depth scan within 3 min at HIMM, information such as the 3D dose distribution, lateral contour images, the Bragg curve, beam spot size, beam position, incident direction, and energy of the beam can be obtained. The Bragg curve acquired from PBDOSE with statistical uncertainties of 1.5% was almost identical to those measured by a commercial water phantom. Besides that, the lateral profiles revealed a remarkable agreement with Gafchromic EBT3 film measurements within differences less than 0.1 mm. PBDOSE shows great prospects in the accurate and efficient measurement of 3D dose distribution for carbon-ion therapy, and the clinical application of this system will improve the efficiency of regular QA procedures to a great extent.

Patient-specific quality assurance in SBRT treatments using 3D polymer gel dosimetry

In stereotactic body radiation therapy (SBRT), patient-specific quality assurance (PSQA) is required due to the complex dose distributions delivered. In this context, polymer gels are considered reliable as a three-dimensional dosimeter. This work used polymer MAGIC-gel sealed in 2 mm thin-wall glass spheres to assess the three-dimensional (3D) dose distribution imparted to patients and compared the results with that provided by the treatment planning system. Four SBRT treatments targeting vertebral, ilium, pelvic lymph node and lung regions were selected for this study. A computed tomography (CT) scan of each glass sphere was obtained, and the patient treatment plan was superimposed on the glass sphere CT image. Thereafter, the spheres were irradiated under similar conditions to the patient using a Varian TrueBeam linear accelerator. A high-spatial resolution optical CT scanner was used to read the glass sphere containing the gel. Planned and delivered 3D dose distributions were compared quantitatively through 3D gamma passing rates at several selected acceptance criteria and qualitatively through the evaluation of dose-profiles, isodose-curves, and dose-volume histograms. We observed that the overall 3D gamma passing rate exceeded 90% at acceptance criteria of 2% dose difference and 2 mm distance-to-agreement for a dose threshold of 20%. The results of this study suggest that the polymer MAGIC-gel sealed in thin-wall glass spheres in conjunction with the optical CT readout method is a suitable PSQA system for SBRT treatments.

A physically constrained Monte Carlo-Neural Network coupling algorithm for BNCT dose calculation.

In boron neutron capture therapy (BNCT)-a form of binary radiotherapy-the primary challenge in treatment planning systems for dose calculations arises from the time-consuming nature of the Monte Carlo (MC) method. Recent progress, including the use of neural networks (NN), has been made to accelerate BNCT dose calculations. However, this approach may result in significant dose errors in both the tumor and the skin, with the latter being a critical organ in BNCT. Furthermore, owing to the lack of physical processes in purely NN-based approaches, their reliability for clinical dose calculations in BNCT is questionable. In this study, a physically constrained MC-NN (PCMC-NN) coupling algorithm is proposed to achieve fast and accurate computation of the BNCT three-dimensional (3D) therapeutic dose distribution. This approach synergizes the high precision of the MC method with the speed of the NN and utilizes physical conservation laws to constrain the coupling process. It addresses the time-consuming issue of the traditional MC method while reducing dose errors. Clinical data were collected from 113 glioblastoma patients. For each patient, the 3D dose distributions for both the coarse and detailed dose grids were calculated using the MC code PHITS. Among these patients, the data from 14 patients were allocated to the test set, 9 to the validation set, and the remaining to the training set. A neural network, 3D-Unet, was built based on the coarse grid dose and patient CT information to enable fast and accurate computation of the 3D detailed grid dose distribution of BNCT. Statistical evaluations, including relative deviation, dose deviation, mean absolute error (MAE), and mean absolute percentage error (MAPE) were conducted. Our findings suggested that the PCMC-NN algorithm substantially outperformed the traditional NN and interpolation methods. Furthermore, the proposed algorithm significantly reduced errors, particularly in the skin and GTV, and improved computational accuracy (hereinafter referred to simply as 'accuracy') with a MAPE range of 1.6%-4.0% and a maximum MAE of 0.3Gy (IsoE) for different organs. The dose-volume histograms generated by the PCMC-NN aligned well with those obtained from the MC method, further validating its accuracy. The PCMC-NN algorithm enhanced the speed and accuracy of BNCT dose calculations by combining the MC method with the NN algorithm. This indicates the significant potential of the proposed algorithm for clinical applications in optimizing treatment planning.

Multi-omics and Multi-VOIs to predict esophageal fistula in esophageal cancer patients treated with radiotherapy

ObjectiveThis study aimed to develop a prediction model for esophageal fistula (EF) in esophageal cancer (EC) patients treated with intensity-modulated radiation therapy (IMRT), by integrating multi-omics features from multiple volumes of interest (VOIs).MethodsWe retrospectively analyzed pretreatment planning computed tomographic (CT) images, three-dimensional dose distributions, and clinical factors of 287 EC patients. Nine groups of features from different combination of omics [Radiomics (R), Dosiomics (D), and RD (the combination of R and D)], and VOIs [esophagus (ESO), gross tumor volume (GTV), and EG (the combination of ESO and GTV)] were extracted and separately selected by unsupervised (analysis of variance (ANOVA) and Pearson correlation test) and supervised (Student T test) approaches. The final model performance was evaluated using five metrics: average area under the receiver-operator-characteristics curve (AUC), accuracy, precision, recall, and F1 score.ResultsFor multi-omics using RD features, the model performance in EG model shows: AUC, 0.817 ± 0.031; 95% CI 0.805, 0.825; p < 0.001, which is better than single VOI (ESO or GTV).ConclusionIntegrating multi-omics features from multi-VOIs enables better prediction of EF in EC patients treated with IMRT. The incorporation of dosiomics features can enhance the model performance of the prediction.

High-fidelity tomographic additive manufacturing for large-volume and high-attenuation situations using expectation maximization algorithm

Tomographic volumetric additive manufacturing (VAM) is an emerging technology that employs patterns projected from all angles surrounding the build volume to tomographically reconstruct a three-dimensional (3D) exposure dose distribution and subsequently photopolymerize the entire target structure. However, the exposure dose distribution reconstructed using the tomographic projection method inherently includes fuzzy boundaries and high background dosages. Consequently, achieving high-fidelity printing using this technique is challenging, particularly when printing large volumes or using high-attenuation materials. The computational algorithm for generating projection patterns plays a significant role in reconstructing high-quality 3D exposure dose distributions and achieving a high-fidelity print. In this study, an expectation-maximization-based method that is more suitable for situations with large volumes or high attenuation is developed. The simulation and printing results reveal that the proposed method can achieve improved reconstruction and high-fidelity printing when printing large volumes or using high-attenuation materials. The proposed method can therefore improve the reliability and robustness of 3D printing technology, particularly in challenging scenarios.

A deep learning-based dose prediction method for evaluation of radiotherapy treatment planning

[Purpose]To address the issue of low accuracy in dose distribution prediction in radiotherapy, we propose a deep learning-based model for predicting three-dimensional dose distribution in tumor radiation therapy. The model utilizes quantitative evaluation methods to assess the treatment plans. [Methods]We selected a dataset of 130 cervical cancer patients, including CT images and target region files. A deep learning U-Net model based on convolutional neural networks and residual blocks was employed to automatically extract multi-scale and multi-level feature maps of CT images, target regions, and anatomical structures of critical organs for intensity-modulated radiation therapy (IMRT) treatment plans and perform three-dimensional dose distribution prediction. Quantitative analysis methods, including error measures such as maximum dose (Dmax), mean dose (Dmean), V20, and D95, were used. [Results]For cervical cancer cases, DVH (dose-volume histogram) graphs were generated based on the evaluation results, directly reflecting the differences between the actual and predicted doses. The actual errors met the basic requirements, and a quantitative evaluation approach was used to optimize the dosimetric parameters. The specific quantification results are: DSC: 86.52 ± 9.31, 95% HD: 3.74 ± 1.49, JD: 0.112 ± 0.026, MSD: 0.067 ± 0.031. [Conclusion]Through training the deep learning model, we have successfully captured the complex nonlinear relationship between IMRT plan feature map parameters and three-dimensional dose distribution. In practical clinical applications, this trained model can accurately predict personalized three-dimensional dose distribution for new patients and effectively assess treatment plans in a quantitative manner. The source code is available at: https://github.com/xiebw9509/Radiotherapy_dose_prediction.

Dosiomics and radiomics improve the prediction of post-radiotherapy neutrophil-lymphocyte ratio in locally advanced non-small cell lung cancer.

To develop and validate a dosiomics and radiomics model based on three-dimensional (3D) dose distribution map and computed tomography (CT) images for the prediction of the post-radiotherapy (post-RT) neutrophil-to-lymphocyte ratio (NLR). This work retrospectively collected 242 locally advanced non-small cell lung cancer (LA-NSCLC) patients who were treated with definitive radiotherapy from 2012 to 2016. The NLR collected one month after the completion of RT was defined as the primary outcome. Clinical characteristics and two-dimensional dosimetric factors calculated from the dose-volume histogram (DVH) were included. A total of 4165 dosiomics and radiomics features were extracted from the 3D dose maps and CT images within five different anatomical regions of interest (ROIs), respectively. Then, a three-step feature selection method was proposed to progressively filter features from coarse to fine: (i) model-based ranking according to individual feature's performance, (ii) maximum relevance and minimum redundancy (mRMR), (iii) select from model based on feature importance calculated with an ensemble of several decision trees. The selected feature subsets were utilized to develop the prediction model with GBDT. All patients were divided into a development set and an independent testing set (2:1). Five-fold cross-validation was applied to the development set for both feature selection and model training procedure. Finally, a fusion model combining dosiomics, radiomics and clinical features was constructed to further improve the prediction results. The area under receiver operating characteristic curve (ROC) were used to evaluate the modelperformance. The clinical-based and DVH-based models showed limited predictive power with AUCs of 0.632 (95% CI: 0.490-0.773) and 0.634 (95% CI: 0.497-0.771), respectively, in the independent testing set. The 9 feature-based dosiomics and 3 feature-based radiomics models showed improved AUCs of 0.738 (95% CI: 0.628-0.849) and 0.689 (95% CI: 0.566-0.813), respectively. The dosiomics & radiomics & clinical fusion model further improved the model's generalization ability with an AUC of 0.765 (95% CI: 0.656-0.874). Dosiomics and radiomics can benefit the prediction of post-RT NLR of LA-NSCLC patients. This can provide a reference for evaluating radiotherapy-relatedinflammation.

Development of DynamicMC for PHITS Monte Carlo package.

Previously, we have developed DynamicMC for modeling relative movement of Oak Ridge National Laboratory phantom in a radiation field for the Monte Carlo N-Particle package (Health Physics. 2023,124(4):301-309). Using this software, three-dimensional dose distributions in a phantom irradiated by a certain mono-energetic (Mono E) source can be deduced through its graphical user interface. In this study, we extended DynamicMC to be used in combination with the Particle and Heavy Ion Transport code System (PHITS) by providing it with a higher flexibility for dynamic movement for an anthropomorphic phantom. For this purpose, we implemented four new functions into the software, which are (1) to generate not only Mono E sources but also those having an energy spectrum of an arbitrary radioisotope (2) to calculate the absorbed doses for several radiologically important organs (3) to automatically average the calculated absorbed doses along the path of the phantom and (4) to generate user-defined slab shielding materials. The first and third items utilize the PHITS-specific modalities named radioisotope-source and sumtally functions, respectively. The computational cost and complexity can be dramatically reduced with these features. We anticipate that the present work and the developed open-source tools will be in the interest of nuclear radiation physics community for research and teaching purposes.

Generalisation of radiotherapy dose calculation for Monte Carlo algorithm combined with 3D Swin-Unet: a multi-institutional IMRT evaluation

Objective. Accurate dose calculations are essential prerequisites for precise radiotherapy. The integration of deep learning into dosimetry could consider computational accuracy and efficiency and has potential applicability to clinical dose calculation. The generalisation of a deep learning dose calculation method (hereinafter referred to as TERMA-Monte Carlo network, T-MC net) was evaluated in clinical practice using intensity-modulated radiotherapy (IMRT) plans for various human body regions and multiple institutions, with the Monte Carlo (MC) algorithm serving as a benchmark. Approach. Sixty IMRT plans were selected from four institutions for testing the head and neck, chest and abdomen, and pelvis regions. Using the MC results as the benchmark, the T-MC net calculation results were used to perform three-dimensional dose distribution and dose-volume histogram (DVH) comparisons of the entire body, planning target volume (PTV) and organs at risk (OARs), respectively, and calculate the mean ±95% confidence interval of gamma pass rate (GPR), percentage of agreement (PA) and dose difference ratio (DDR) of dose indices D95, D50, and D5. Main results. For the entire body, the GPRs of 3%/3 mm, 2%/2 mm, 2%/1 mm, and the PA were 99.62 ± 0.32%, 98.50 ± 1.09%, 95.60 ± 2.90% and 97.80 ± 1.12%, respectively. For the PTV, the GPRs of 3%/3 mm, 2%/2 mm, 2%/1 mm and the PA were 98.90 ± 1.00%, 95.78 ± 2.83%, 92.23 ± 4.74% and 98.93 ± 0.62%, respectively. The absolute value of average DDR was less than 1.4%. Significance. We proposed a general dose calculation framework based on deep learning, using the MC algorithm as a benchmark, performing a generalisation test for IMRT treatment plans across multiple institutions. The framework provides high computational speed while maintaining the accuracy of MC and may become an effective dose algorithm engine in treatment planning, adaptive radiotherapy, and dose verification.

Deep-learning Method for the Prediction of Three-Dimensional Dose Distribution for Left Breast Cancer Conformal Radiation Therapy

AimsAn increase in the demand of a new generation of radiotherapy planning systems based on learning approaches has been reported. At this stage, the new approach is able to improve the planning speed while saving a reasonable level of plan quality, compared with available planning systems. We believe that new achievements, such as deep-learning models, will be able to review the issue from a different point of view. Materials and methodsThe data of 120 breast cancer patients were used to train and test the three-dimensional U-Res-Net model. The network input was computed tomography images and patients' contouring, while the patients' dose distribution was addressed as the output of the model proposed. The predicted dose distributions, created by the model for 10 test patients, were then compared with corresponding dose distributions calculated by a reliable treatment planning system. In particular, the dice similarity coefficients for different isodose volumes, dose difference and mean absolute errors (MAE) for all voxels inside the body, Dmean, D98%, D50%, D2%, V95% for planning target volume and organs at risk were calculated and were statistically analysed with the paired-samples t-test. ResultsThe average dose difference for all patients and voxels in body was 0.60 ± 2.81%. The MAE varied from 3.85 ± 6.65% to 8.06 ± 10.00%. The average MAE for test cases was 5.71 ± 1.19%. The average dice similarity coefficients for isodose volumes was 0.91 ± 0.03. The three-dimensional gamma passing rates with 3 mm/3% criteria varied from 78.99% to 97.58% for planning target volume and organs at risk, respectively. ConclusionsThe investigation showed that a deep-learning model can be applied to predict the three-dimensional dose distribution with optimal accuracy and precision for patients with left breast cancer. As further study, the model can be extended to predict dose distribution in other cancers.

Deep learning-based dose prediction to improve the plan quality of volumetric modulated arc therapy for gynecologic cancers.

In recent years, deep-learning models have been used to predict entire three-dimensional dose distributions. However, the usability of dose predictions to improve plan quality should be further investigated. To develop a deep-learning model to predict high-quality dose distributions for volumetric modulated arc therapy (VMAT) plans for patients with gynecologic cancer and to evaluate their usability in driving plan quality improvements. A total of 79 VMAT plans for the female pelvis were used to train (47 plans), validate (16 plans), and test (16 plans) 3D dense dilated U-Net models to predict 3D dose distributions. The models received the normalized CT scan, dose prescription, and target and normal tissue contours as inputs. Three models were used to predict the dose distributions for plans in the test set. A radiation oncologist specializing in the treatment of gynecologic cancers scored the test set predictions using a 5-point scale (5, acceptable as-is; 4, prefer minor edits; 3, minor edits needed; 2, major edits needed; and 1, unacceptable). The clinical plans for which the dose predictions indicated that improvements could be made were reoptimized with constraints extracted from the predictions. The predicted dose distributions in the test set were of comparable quality to the clinical plans. The mean voxel-wise dose difference was -0.14±0.46Gy. The percentage dose differences in the predicted target metrics of and were -1.05%±0.59% and 0.21%±0.28%, respectively. The dose differences in the predicted organ at risk mean and maximum doses were -0.30±1.66Gy and -0.42±2.07Gy, respectively. A radiation oncologist deemed all of the predicted dose distributions clinically acceptable; 12 received a score of 5, and four received a score of 4. Replanning of flagged plans (five plans) showed that the original plans could be further optimized to give dose distributions close to the predicted dose distributions. Deep-learning dose prediction can be used to predict high-quality and clinically acceptable dose distributions for VMAT female pelvis plans, which can then be used to identify plans that can be improved with additional optimization.

Enhanced prediction of postoperative radiotherapy-induced esophagitis in non-small cell lung cancer: Dosiomic model development in a real-world cohort and validation in the PORT-C randomized controlled trial.

Radiotherapy-induced esophagitis (RE) diminishes the quality of life and interrupts treatment in patients with non-small cell lung cancer (NSCLC) undergoing postoperative radiotherapy. Dosimetric models showed limited capability in predicting RE. We aimed to develop dosiomic models to predict RE. Models were trained with a real-world cohort and validated with PORT-C randomized controlled trial cohort. Patients with NSCLC undergoing resection followed by postoperative radiotherapy between 2004 and 2015 were enrolled. The endpoint was grade ≥2 RE. Esophageal three-dimensional dose distribution features were extracted using handcrafted and convolutional neural network (CNN) methods, screened using an entropy-based method, and selected using minimum redundancy and maximum relevance. Prediction models were built using logistic regression. The areas under the receiver operating characteristic curve (AUC) and precision-recall curve were used to evaluate prediction model performance. A dosimetric model was built for comparison. A total of 190 and 103 patients were enrolled in the training and validation sets, respectively. Using handcrafted and CNN methods, 107 and 4096 features were derived, respectively. Three handcrafted, four CNN-extracted and three dosimetric features were selected. AUCs of training and validation sets were 0.737 and 0.655 for the dosimetric features, 0.730 and 0.724 for handcrafted features, and 0.812 and 0.785 for CNN-extracted features, respectively. Precision-recall curves revealed that CNN-extracted features outperformed dosimetric and handcrafted features. Prediction models may identify patients at high risk of developing RE. Dosiomic models outperformed the dosimetric-feature model in predicting RE. CNN-extracted features were more predictive but less interpretable than handcrafted features.

A comparative study of deep learning-based knowledge-based planning methods for 3D dose distribution prediction of head and neck.

In this paper, we compare four novel knowledge-based planning (KBP) algorithms using deep learning to predict three-dimensional (3D) dose distributions of head and neck plans using the same patients' dataset and quantitative assessment metrics. A dataset of 340 oropharyngeal cancer patients treated with intensity-modulated radiation therapy was used in this study, which represents the AAPM OpenKBP - 2020 Grand Challenge dataset. Four 3D convolutional neural network architectures were built. The models were trained on 64% of the data set and validated on 16% for voxel-wise dose predictions: U-Net, attention U-Net, residual U-Net (Res U-Net), and attention Res U-Net. The trained models were then evaluated for their performance on a test data set (20% of the data) by comparing the predicted dose distributions against the ground-truth using dose statistics and dose-volume indices. The four KBP dose prediction models exhibited promising performance with an averaged mean absolute dose error within the body contour <3Gy on 68 plans in the test set. The average difference in predicting the D99 index for all targets was 0.92Gy (p=0.51) for attention Res U-Net, 0.94Gy (p=0.40) for Res U-Net, 2.94Gy (p=0.09) for attention U-Net, and 3.51Gy (p=0.08) for U-Net. For the OARs, the values for the and indices were 2.72Gy (p<0.01) for attention Res U-Net, 2.94Gy (p<0.01) for Res U-Net, 1.10Gy (p<0.01) for attention U-Net, 0.84Gy (p<0.29) for U-Net. All models demonstrated almost comparable performance for voxel-wise dose prediction. KBP models that employ 3D U-Net architecture as a base could be deployed for clinical use to improve cancer patient treatment by creating plans with consistent quality and making the radiotherapy workflow more efficient.

Rapid estimation of patient-specific organ doses using a deep learning network.

Patient-specific organ-dose estimation in diagnostic CT examinations can provide useful insights on individualized secondary cancer risks, protocol optimization and patient management. Current dose estimation techniques mainly rely on time-consuming Monte Carlo methods or/and generalized anthropomorphic phantoms. We proposed a proof-of-concept rapid workflow based on deep learning networks to estimate organ doses for individuals following thorax Computed Tomography (CT) examinations METHODS: CT scan data from 95 individuals undergoing thorax CT examinations were used. Monte Carlo simulations were performed and three-dimensional (3D) dose distributions for each patient were obtained. A fully connected sequential deep learning network model was constructed and trained for each organ considered in this study. Water-equivalent diameter (WED), scan length and tube current were the independent variables. Organ doses for heart, lungs, esophagus, and bones were calculated from the Monte Carlo 3D distribution and used to train the deep learning networks. Organ dose predictions from each network were evaluated using an independent data set of 19 patients. The trained networks provided organ dose predictions within a second. There was very good agreement between the deep learning network predictions and reference organ dose values calculated from Monte Carlo simulations. The average difference was -1.5% for heart, -1.6% for esophagus, -1.0% for lungs, and -0.4% for bones in the 95 patients dataset, and -5.1%, 4.3%, 0.9%, and 1.4% respectively in the 19 patients test dataset. The proposed workflow demonstrated that patient specific organ-doses can be estimated in nearly real-time using deep learning networks. The workflow can be readily implemented and requires a small set of representative data for training. This article is protected by copyright. All rights reserved.

Multi-omics to predict acute radiation esophagitis in patients with lung cancer treated with intensity-modulated radiation therapy

PurposeThe study aimed to predict acute radiation esophagitis (ARE) with grade ≥ 2 for patients with locally advanced lung cancer (LALC) treated with intensity-modulated radiation therapy (IMRT) using multi-omics features, including radiomics and dosiomics.Methods161 patients with stage IIIA−IIIB LALC who received chemoradiotherapy (CRT) or radiotherapy by IMRT with a prescribed dose from 45 to 70 Gy from 2015 to 2019 were enrolled retrospectively. All the toxicity gradings were given following the Common Terminology Criteria for Adverse Events V4.0. Multi-omics features, including radiomics, dosiomics (including dose−volume histogram dosimetric parameters), were extracted based on the planning CT image and three-dimensional dose distribution. All data were randomly divided into training cohorts (N = 107) and testing cohorts (N = 54). In the training cohorts, features with reliably high outcome relevance and low redundancy were selected under random patient subsampling. Four classification models (using clinical factors (CF) only, using radiomics features (RFs) only, dosiomics features (DFs) only, and the hybrid features (HFs) containing clinical factors, radiomics and dosiomics) were constructed employing the Ridge classifier using two-thirds of randomly selected patients as the training cohort. The remaining patient was treated as the testing cohort. A series of models were built with 30 times training–testing splits. Their performances were assessed using the area under the ROC curve (AUC) and accuracy.ResultsAmong all patients, 51 developed ARE grade ≥ 2, with an incidence of 31.7%. Next, 8990 radiomics and 213 dosiomics features were extracted, and 3, 6, 12, and 13 features remained after feature selection in the CF, DF, RF and DF models, respectively. The RF and HF models achieved similar classification performance, with the training and testing AUCs of 0.796 ± 0.023 (95% confidence interval (CI [0.79, 0.80])/0.744 ± 0.044 (95% CI [0.73, 0.76]) and 0.801 ± 0.022 (95% CI [0.79, 0.81]) (p = 0.74), respectively. The model performances using CF and DF features were poorer, with training and testing AUCs of 0.573 ± 0.026 (95% CI [0.56, 0.58])/ 0.509 ± 0.072 (95% CI [0.48, 0.53]) and 0.679 ± 0.027 (95% CI [0.67, 0.69])/0.604 ± 0.041 (95% CI [0.53, 0.63]) compared with the above two models (p < 0.001), respectively.ConclusionsIn LALC patients treated with CRT IMRT, the ARE grade ≥ 2 can be predicted using the pretreatment radiotherapy image features. To predict ARE, the multi-omics features had similar predictability with radiomics features; however, the dosiomics features and clinical factors had a limited classification performance.

End-to-End QA with Polymer Gel Dosimeter for Photon Beam Radiation Therapy.

With the complexity and high demands on quality assurance (QA) of photon beam radiation therapy, end-to-end (E2E) QA is necessary to validate the entire treatment workflow from pre-treatment imaging to beam delivery. A polymer gel dosimeter is a promising tool for three-dimensional (3D) dose distribution measurement. The purpose of this study is to design a fast "one delivery" polymethyl methacrylate (PMMA) phantom with a polymer gel dosimeter for the E2E QA test of the photon beam. The one delivery phantom is composed of ten calibration cuvettes for the calibration curve measurement, two 10 cm gel dosimeter inserts for the dose distribution measurement, and three 5.5 cm gel dosimeters for the square field measurement. The one delivery phantom holder is comparable in size and shape to that of a human thorax and abdomen. In addition, an anthropomorphic head phantom was employed to measure the patient-specific dose distribution of a VMAT plan. The E2E dosimetry was verified by undertaking the whole RT procedure (immobilization, CT simulation, treatment planning, phantom set-up, imaged-guided registration, and beam delivery). The calibration curve, field size, and patient-specific dose were measured with a polymer gel dosimeter. The positioning error can be mitigated with the one-delivery PMMA phantom holder. The delivered dose measured with a polymer gel dosimeter was compared with the planned dose. The gamma passing rate is 86.64% with the MAGAT-f gel dosimeter. The results ascertain the feasibility of the one delivery phantom with a polymer gel dosimeter for a photon beam in E2E QA. The QA time can be reduced with the designed one delivery phantom.