Three‐dimensional dose prediction for lung IMRT patients with deep neural networks: robust learning from heterogeneous beam configurations

Dose prediction for cervical cancer in radiotherapy based on the beam channel generative adversarial network

Background

Existing deep learning methods, such as generative adversarial network (GAN) technology, face challenges when dealing with mixed datasets, which involve a combination of Intensity Modulated Radiotherapy (IMRT) and Volumetric Modulated Arc Therapy (VMAT). This issue significantly complicates the application of dose prediction in the field of radiotherapy. In this study, we propose a novel approach called beam channel GAN (Bc-GAN) to address the task of radiation dose prediction for mixed datasets. Bc-GAN introduces a dose prediction calculation method that requires less precision. By defining an approximate range for dose prediction, Bc-GAN limits the physical range of GAN prediction, resulting in more reasonable dose distribution predictions.

Methods

We adopt a beam angle weighting method to determine the beam angle in the dose calculation. The dose of the beam with the highest weight is calculated using medical images and is then inputted into the artificial intelligence dose prediction model as the input channel. Additionally, we collect data from a total of 346 patients with Cervical Cancer (CC) for dataset. After cleaning the data, we exclude 51 cases with incomplete organ delineation, leaving us with 295 cases (IMRT: VMAT = 137:158) randomly divided into three sets: the training set, the validation set, and the test set, with proportions of 205:60:30, respectively. The assessment of model predictions was conducted via an analysis of dose distributions on the tomographic plane, dose volume histogram (DVH), and dosimetric parameters within the target zones and organs at risk (OAR).

Results

After DVH analysis, minimal discrepancy was found between predicted and actual dose distributions in PTV and OAR. The predicted distribution aligned with clinical standards. Dosimetric parameters for PTV were generally lower in the predicted model, except for homogeneity index (HI) (0.238 ± 0.024, P = 0.017) and Dmax (53.599 ± 0.710 Gy, P = 1.8e-05). The prediction model varied in estimating doses for six organs. Specifically, small intestine showed higher V20 (67.92 ± 51.64 %, P = 0.019) and V30 (57.171 ± 1.213 %, P = 0.024) than manual planning. A similar trend was seen in colon's V30 (37.13 ± 61.14 %, P = 0.016). However, predicted bladder V30 (87.51 ± 41.44 %, P = 2.03e-16) was lower, indicating significant dosimetric differences.

Conclusion

Overall, this study presents an innovative prediction method for CC in radiotherapy using the Bc-GAN model, addressing the challenges posed by different radiotherapy techniques. The proposed approach allows IMRT and VMAT in radiotherapy to be used as training sets, enabling the potential for large-scale engineering and commercialization applications of artificial intelligence (AI). The Bc-GAN-based prediction method for CC in radiotherapy not only reduces the amount of data needed for the training set but also expedites the model generation process. This approach can be applied to guide the development of clinical radiation therapy plans. Furthermore, future studies should consider extending the dose prediction method to encompass other types of tumors.

Artificial intelligence (AI) applications in improvement of IMRT and VMAT radiotherapy treatment planning processes: A systematic review

INTRODUCTION

Radiotherapy is a common option in the treatment of many types of cancer. Intensity-Modulated Radiation Therapy (IMRT) and Volumetric-Modulated Arc Therapy (VMAT) are the latest radiotherapy techniques. However, clinicians face problems due to these techniques' complexity and time-consuming planning. Various studies have pointed out the importance and role of artificial intelligence (AI) in radiotherapy and accelerating and improving its quality. This research explores different AI methods in different fields of IMRT and VMAT. This study evaluated both quantitative and qualitative methods used within the reviewed articles.

METHODS

Various articles were reviewed from Google Scholar, Science Direct, and PubMed databases between 2018 and 2024. According to PRISMA 2020 guidelines, study selection processes, screening, and inclusion and exclusion criteria were defined. The critical Appraisal Skill Program qualitative checklist tool was used for the qualitative evaluation of articles.

RESULTS

26 articles met the inclusion among the 33 articles obtained. The search procedure was displayed using the PRISMA flow diagram. The evaluation of the articles shows the automation of various treatment planning processes by AI methods and their better performance than traditional methods. The qualitative evaluation of studies has demonstrated the high quality of all studies. The lowest score obtained from the qualitative evaluation of the article is 7 out of 9.

CONCLUSION

AI methods used in radiotherapy reduce time and increase prediction accuracy. They also work better than other methods in different areas, such as dose prediction, treatment design, and dose delivery.

IMPLICATIONS FOR PRACTICE

Healthcare providers should consider integrating artificial intelligence technologies into their practice to optimize treatment planning and enhance patient care in radiation therapy. Additionally, fostering collaboration between radiotherapy experts and artificial intelligence specialists can significantly improve the development and application of AI technologies in this field.

Deep evidential learning for radiotherapy dose prediction

BACKGROUND

As we navigate towards integrating deep learning methods in the real clinic, a safety concern lies in whether and how the model can express its own uncertainty when making predictions. In this work, we present a novel application of an uncertainty-quantification framework called Deep Evidential Learning in the domain of radiotherapy dose prediction.

METHOD

Using medical images of the Open Knowledge-Based Planning Challenge dataset, we found that this model can be effectively harnessed to yield uncertainty estimates that inherited correlations with prediction errors upon completion of network training. This was achieved only after reformulating the original loss function for a stable implementation.

RESULTS

We found that (i) epistemic uncertainty was highly correlated with prediction errors, with various association indices comparable or stronger than those for Monte-Carlo Dropout and Deep Ensemble methods, (ii) the median error varied with uncertainty threshold much more linearly for epistemic uncertainty in Deep Evidential Learning relative to these other two conventional frameworks, indicative of a more uniformly calibrated sensitivity to model errors, (iii) relative to epistemic uncertainty, aleatoric uncertainty demonstrated a more significant shift in its distribution in response to Gaussian noise added to CT intensity, compatible with its interpretation as reflecting data noise.

CONCLUSION

Collectively, our results suggest that Deep Evidential Learning is a promising approach that can endow deep-learning models in radiotherapy dose prediction with statistical robustness. We have also demonstrated how this framework leads to uncertainty heatmaps that correlate strongly with model errors, and how it can be used to equip the predicted Dose-Volume-Histograms with confidence intervals.

Precision dose prediction for breast cancer patients undergoing IMRT: The Swin-UMamba-Channel Model

BACKGROUND

Radiation therapy is one of the crucial treatment modalities for cancer. An excellent radiation therapy plan relies heavily on an outstanding dose distribution map, which is traditionally generated through repeated trials and adjustments by experienced physicists. However, this process is both time-consuming and labor-intensive, and it comes with a degree of subjectivity. Now, with the powerful capabilities of deep learning, we are able to predict dose distribution maps more accurately, effectively overcoming these challenges.

METHODS

In this study, we propose a novel Swin-UMamba-Channel prediction model specifically designed for predicting the dose distribution of patients with left breast cancer undergoing radiotherapy after total mastectomy. This model integrates anatomical position information of organs and ray angle information, significantly enhancing prediction accuracy. Through iterative training of the generator (Swin-UMamba) and discriminator, the model can generate images that closely match the actual dose, assisting physicists in quickly creating DVH curves and shortening the treatment planning cycle. Our model exhibits excellent performance in terms of prediction accuracy, computational efficiency, and practicality, and its effectiveness has been further verified through comparative experiments with similar networks.

RESULTS

The results of the study indicate that our model can accurately predict the clinical dose of breast cancer patients undergoing intensity-modulated radiation therapy (IMRT). The predicted dose range is from 0 to 50 Gy, and compared with actual data, it shows a high accuracy with an average Dice similarity coefficient of 0.86. Specifically, the average dose change rate for the planning target volume ranges from 0.28 % to 1.515 %, while the average dose change rates for the right and left lungs are 2.113 % and 0.508 %, respectively. Notably, due to their small sizes, the heart and spinal cord exhibit relatively higher average dose change rates, reaching 3.208 % and 1.490 %, respectively. In comparison with similar dose studies, our model demonstrates superior performance. Additionally, our model possesses fewer parameters, lower computational complexity, and shorter processing time, further enhancing its practicality and efficiency. These findings provide strong evidence for the accuracy and reliability of our model in predicting doses, offering significant technical support for IMRT in breast cancer patients.

CONCLUSION

This study presents a novel Swin-UMamba-Channel dose prediction model, and its results demonstrate its precise prediction of clinical doses for the target area of left breast cancer patients undergoing total mastectomy and IMRT. These remarkable achievements provide valuable reference data for subsequent plan optimization and quality control, paving a new path for the application of deep learning in the field of radiation therapy.

Predicting the 3-Dimensional Dose Distribution of Multilesion Lung Stereotactic Ablative Radiation Therapy With Generative Adversarial Networks

PURPOSE

Because SABR therapy is being used to treat greater numbers of lung metastases, selecting the optimal dose and fractionation to balance local failure and treatment toxicity becomes increasingly challenging. Multilesion lung SABR therapy plans include spatially diverse lesions with heterogeneous prescriptions and interacting dose distributions. In this study, we developed and evaluated a generative adversarial network (GAN) to provide real-time dosimetry predictions for these complex cases.

METHODS AND MATERIALS

A GAN was trained to predict dosimetry on a data set of patients who received SABR therapy for lung lesions at a tertiary center. Model input included the planning computed tomography scan, the organs at risk (OARs) and target structures, and an initial estimate of exponential dose fall-off. Multilesion plans were split 80/20 for training and evaluation. Models were evaluated on voxel-voxel, clinical dose-volume histogram, and conformality metrics. An out-of-sample validation and analysis of model variance were performed.

RESULTS

There were 125 multilesion plans from 102 patients with 357 lesions. Patients were treated for 2 to 7 lesions, with 19 unique dose-fractionation schemes over 1 to 3 courses of treatment. The out-of-sample validation set contained an additional 90 plans from 80 patients. The mean absolute difference and gamma pass fraction between the predicted and true dosimetry was <3 Gy and >90% for all OARs. The absolute differences in lung V20 and CV14 were 1.40% ± 0.99% and 75.8 ± 42.0 cc, respectively. The ratios of predicted to true R50%, R100%, and D2cm were 1.00 ± 0.16, 0.96 ± 0.32, and 1.01 ± 0.36, respectively. The out-of-sample validation set maintained mean absolute difference and gamma pass fraction of <3 Gy and >90%, respectively for all OARs. The median standard deviation of variance in V20 and CV14 prediction was 0.49% and 22.2 cc, respectively.

CONCLUSIONS

A GAN for predicting the 3-D dosimetry of complex multilesion lung SABR therapy is presented. Rapid dosimetry prediction can be used to assess treatment feasibility and explore dosimetric differences between varying prescriptions.

Clinical implementation and evaluation of deep learning-assisted automatic radiotherapy treatment planning for lung cancer

PURPOSE

The purpose of the study is to investigate the clinical application of deep learning (DL)-assisted automatic radiotherapy planning for lung cancer.

METHODS

A DL model was developed for predicting patient-specific doses, trained and validated on a dataset of 235 patients with diverse target volumes and prescriptions. The model was integrated into clinical workflow with DL-predicted objective functions. The automatic plans were retrospectively designed for additional 50 treated manual volumetric modulated arc therapy (VMAT) plans. A comparison was made between automatic and manual plans in terms of dosimetric indexes, monitor units (MUs) and planning time. Plan quality metric (PQM) encompassing these indexes was evaluated, with higher PQM values indicating superior plan quality. Qualitative evaluations of two plans were conducted by four reviewers.

RESULTS

The PQM score was 40.7 ± 13.1 for manual plans and 40.8 ± 13.5 for automatic plans (P = 0.75). Compared to manual plans, the targets coverage and homogeneity of automatic plans demonstrated no significant difference. Manual plans exhibited better sparing for lung in V5 (difference: 1.8 ± 4.2 %, P = 0.02), whereas automatic plans showed enhanced sparing for heart in V30 (difference: 1.4 ± 4.7 %, P = 0.02) and for spinal cord in Dmax (difference: 0.7 ± 4.7 Gy, P = 0.04). The planning time and MUs of automatic plans were significantly reduced by 70.5 ± 20.0 min and 97.4 ± 82.1. Automatic plans were deemed acceptable in 88 % of the reviews (176/200).

CONCLUSIONS

The DL-assisted approach for lung cancer notably decreased planning time and MUs, while demonstrating comparable or superior quality relative to manual plans. It has the potential to provide benefit to lung cancer patients.

Deep learning applied to dose prediction in external radiation therapy: A narrative review

Over the last decades, the use of artificial intelligence, machine learning and deep learning in medical fields has skyrocketed. Well known for their results in segmentation, motion management and posttreatment outcome tasks, investigations of machine learning and deep learning models as fast dose calculation or quality assurance tools have been present since 2000. The main motivation for this increasing research and interest in artificial intelligence, machine learning and deep learning is the enhancement of treatment workflows, specifically dosimetry and quality assurance accuracy and time points, which remain important time-consuming aspects of clinical patient management. Since 2014, the evolution of models and architectures for dose calculation has been related to innovations and interest in the theory of information research with pronounced improvements in architecture design. The use of knowledge-based approaches to patient-specific methods has also considerably improved the accuracy of dose predictions. This paper covers the state of all known deep learning architectures and models applied to external radiotherapy with a description of each architecture, followed by a discussion on the performance and future of deep learning predictive models in external radiotherapy.

Prediction of dose distributions for non-small cell lung cancer patients using MHA-ResUNet

BACKGROUND

The current level of automation in the production of radiotherapy plans for lung cancer patients is relatively low. With the development of artificial intelligence, it has become a reality to use neural networks to predict dose distributions and provide assistance for radiation therapy planning. However, due to the significant individual variability in the distribution of non-small cell lung cancer (NSCLC) planning target volume (PTV) and the complex spatial relationships between the PTV and organs at risk (OARs), there is still a lack of a high-precision dose prediction network tailored to the characteristics of NSCLC.

PURPOSE

To assist in the development of volumetric modulated arc therapy (VMAT) plans for non-small cell lung cancer patients, a deep neural network is proposed to predict high-precision dose distribution.

METHODS

This study has developed a network called MHA-ResUNet, which combines a large-kernel dilated convolution module and multi-head attention (MHA) modules. The network was trained based on 80 VMAT plans of NSCLC patients. CT images, PTV, and OARs were fed into the independent input channel. The dose distribution was taken as the output to train the model. The performance of this network was compared with that of several commonly used networks, and the networks' performance was evaluated based on the voxel-level mean absolute error (MAE) within the PTV and OARs, as well as the error in clinical dose-volume metrics.

RESULTS

The MAE between the predicted dose distribution and the manually planned dose distribution within the PTV is 1.43 Gy, and the D95 error is less than 1 Gy. Compared with the other three commonly used networks, the dose error of the MHA-ResUNet is the smallest in PTV and OARs.

CONCLUSIONS

The proposed MHA-ResUNet network improves the receptive field and filters the shallow features to learn the relative spatial relation between the PTV and the OARs, enabling accurate prediction of dose distributions in NSCLC patients undergoing VMAT radiotherapy.

Can input reconstruction be used to directly estimate uncertainty of a dose prediction U-Net model?

BACKGROUND

The reliable and efficient estimation of uncertainty in artificial intelligence (AI) models poses an ongoing challenge in many fields such as radiation therapy. AI models are intended to automate manual steps involved in the treatment planning workflow. We focus in this study on dose prediction models that predict an optimal dose trade-off for each new patient for a specific treatment modality. They can guide physicians in the optimization, be part of automatic treatment plan generation or support decision in treatment indication. Most common uncertainty estimation methods are based on Bayesian approximations, like Monte Carlo dropout (MCDO) or Deep ensembling (DE). These two techniques, however, have a high inference time (i.e., require multiple inference passes) and might not work for detecting out-of-distribution (OOD) data (i.e., overlapping uncertainty estimate for in-distribution (ID) and OOD).

PURPOSE

In this study, we present a direct uncertainty estimation method and apply it for a dose prediction U-Net architecture. It can be used to flag OOD data and give information on the quality of the dose prediction.

METHODS

Our method consists in the addition of a branch decoding from the bottleneck which reconstructs the CT scan given as input. The input reconstruction error can be used as a surrogate of the model uncertainty. For the proof-of-concept, our method is applied to proton therapy dose prediction in head and neck cancer patients. A dataset of 60 oropharyngeal patients was used to train the network using a nested cross-validation approach with 11 folds (training: 50 patients, validation: 5 patients, test: 5 patients). For the OOD experiment, we used 10 extra patients with a different head and neck sub-location. Accuracy, time-gain, and OOD detection are analyzed for our method in this particular application and compared with the popular MCDO and DE.

RESULTS

The additional branch did not reduce the accuracy of the dose prediction model. The median absolute error is close to zero for the target volumes and less than 1% of the dose prescription for organs at risk. Our input reconstruction method showed a higher Pearson correlation coefficient with the prediction error (0.620) than DE (0.447) and MCDO (between 0.599 and 0.612). Moreover, our method allows an easier identification of OOD (no overlap for ID and OOD data and a Z-score of 34.05). The uncertainty is estimated simultaneously to the regression task, therefore requires less time and computational resources.

CONCLUSIONS

This study shows that the error in the CT scan reconstruction can be used as a surrogate of the uncertainty of the model. The Pearson correlation coefficient with the dose prediction error is slightly higher than state-of-the-art techniques. OOD data can be more easily detected and the uncertainty metric is computed simultaneously to the regression task, therefore faster than MCDO or DE. The code and pretrained model are available on the gitlab repository: https://gitlab.com/ai4miro/ct-reconstruction-for-uncertainty-quatification-of-hdunet.

Application and progress of artificial intelligence in radiation therapy dose prediction

Radiation therapy (RT) nowadays is a main treatment modality of cancer. To ensure the therapeutic efficacy of patients, accurate dose distribution is often required, which is a time-consuming and labor-intensive process. In addition, due to the differences in knowledge and experience among participants and diverse institutions, the predicted dose are often inconsistent. In last several decades, artificial intelligence (AI) has been applied in various aspects of RT, several products have been implemented in clinical practice and confirmed superiority. In this paper, we will review the research of AI in dose prediction, focusing on the progress in deep learning (DL).

AS-NeSt: A Novel 3D Deep Learning Model for Radiation Therapy Dose Distribution Prediction in Esophageal Cancer Treatment With Multiple Prescriptions

PURPOSE

Implementing artificial intelligence technologies allows for the accurate prediction of radiation therapy dose distributions, enhancing treatment planning efficiency. However, esophageal cancers present unique challenges because of tumor complexity and diverse prescription types. Additionally, limited data availability hampers the effectiveness of existing artificial intelligence models. This study developed a deep learning model, trained on a diverse data set of esophageal cancer prescriptions, to improve dose prediction accuracy.

METHODS AND MATERIALS

We retrospectively collected data from 530 patients with esophageal cancer, including single-target and simultaneous integrated boost prescriptions, for model building. The proposed Asymmetric ResNeSt (AS-NeSt) model features novel 3-dimensional (3D) ResNeSt blocks and an asymmetrical architecture. We constructed a loss function targeting global and local doses and validated the model's performance against existing alternatives. Model-assisted experiments were used to validate its clinical benefits.

RESULTS

The AS-NeSt model maintained an absolute prediction error below 5% for each dosimetric metric. The average Dice similarity coefficient for isodose volumes was 0.93. The model achieved an average relative prediction error of 2.02%, statistically lower than Hierarchically Densely Connected U-net (4.17%), DoseNet (2.35%), and Densely Connected Network (3.65%). It also demonstrated significantly fewer parameters and shorter prediction times. Clinically, the AS-NeSt model raised physicians' ability to accurately preassess appropriate treatment methods before planning from 95.24% to 100%, reduced planning time by over 61% for junior dosimetrists and 52% for senior dosimetrists, and decreased both inter- and intra-dosimetrist discrepancies by more than 50%.

CONCLUSIONS

The AS-NeSt model, developed with innovative 3D ResNeSt blocks and an asymmetrical encoder-decoder structure, has been validated using clinical esophageal cancer patient data. It accurately predicts 3D dose distributions for various prescriptions, including simultaneous integrated boost, showing potential to improve the management of esophageal cancer treatment in a clinical setting.

A generalization performance study on the boosting radiotherapy dose calculation engine based on super-resolution

PURPOSE

During the radiation treatment planning process, one of the time-consuming procedures is the final high-resolution dose calculation, which obstacles the wide application of the emerging online adaptive radiotherapy techniques (OLART). There is an urgent desire for highly accurate and efficient dose calculation methods. This study aims to develop a dose super resolution-based deep learning model for fast and accurate dose prediction in clinical practice.

METHOD

A Multi-stage Dose Super-Resolution Network (MDSR Net) architecture with sparse masks module and multi-stage progressive dose distribution restoration method were developed to predict high-resolution dose distribution using low-resolution data. A total of 340 VMAT plans from different disease sites were used, among which 240 randomly selected nasopharyngeal, lung, and cervix cases were used for model training, and the remaining 60 cases from the same sites for model benchmark testing, and additional 40 cases from the unseen site (breast and rectum) was used for model generalizability evaluation. The clinical calculated dose with a grid size of 2 mm was used as baseline dose distribution. The input included the dose distribution with 4 mm grid size and CT images. The model performance was compared with HD U-Net and cubic interpolation methods using Dose-volume histograms (DVH) metrics and global gamma analysis with 1%/1 mm and 10% low dose threshold. The correlation between the prediction error and the dose, dose gradient, and CT values was also evaluated.

RESULTS

The prediction errors of MDSR were 0.06-0.84% of Dmean indices, and the gamma passing rate was 83.1-91.0% on the benchmark testing dataset, and 0.02-1.03% and 71.3-90.3% for the generalization dataset respectively. The model performance was significantly higher than the HD U-Net and interpolation methods (p < 0.05). The mean errors of the MDSR model decreased (monotonously by 0.03-0.004%) with dose and increased (by 0.01-0.73%) with the dose gradient. There was no correlation between prediction errors and the CT values.

CONCLUSION

The proposed MDSR model achieved good agreement with the baseline high-resolution dose distribution, with small prediction errors for DVH indices and high gamma passing rate for both seen and unseen sites, indicating a robust and generalizable dose prediction model. The model can provide fast and accurate high-resolution dose distribution for clinical dose calculation, particularly for the routine practice of OLART.

Deep-neural network approaches for predicting 3D dose distribution in intensity-modulated radiotherapy of the brain tumors

PURPOSE

The aim of this study is to reduce treatment planning time by predicting the intensity-modulated radiotherapy 3D dose distribution using deep learning for brain cancer patients. "For this purpose, two different approaches in dose prediction, i.e., first only planning target volume (PTV) and second PTV with organs at risk (OARs) as input of the U-net model, are employed and their results are compared."

METHODS AND MATERIALS

The data of 99 patients with glioma tumors referred for IMRT treatment were used so that the images of 90 patients were regarded as training datasets and the others were for the test. All patients were manually planned and treated with sixth-field IMRT; the photon energy was 6MV. The treatment plans were done with the Collapsed Cone Convolution algorithm to deliver 60 Gy in 30 fractions.

RESULTS

The obtained accuracy and similarity for the proposed methods in dose prediction when compared to the clinical dose distributions on test patients according to MSE, dice metric and SSIM for the Only-PTV and PTV-OARs methods are on average (0.05, 0.851, 0.83) and (0.056, 0.842, 0.82) respectively. Also, dose prediction is done in an extremely short time.

CONCLUSION

The same results of the two proposed methods prove that the presence of OARs in addition to PTV does not provide new knowledge to the network and only by defining the PTV and its location in the imaging slices, does the dose distribution become predictable. Therefore, the Only-PTV method by eliminating the process of introducing OARs can reduce the overall designing time of treatment by IMRT in patients with glioma tumors.

Predicting voxel-level dose distributions of single-isocenter volumetric modulated arc therapy treatment plan for multiple brain metastases

Purpose

Brain metastasis is a common, life-threatening neurological problem for patients with cancer. Single-isocenter volumetric modulated arc therapy (VMAT) has been popularly used due to its highly conformal dose and short treatment time. Accurate prediction of its dose distribution can provide a general standard for evaluating the quality of treatment plan. In this study, a deep learning model is applied to the dose prediction of a single-isocenter VMAT treatment plan for radiotherapy of multiple brain metastases.

Method

A U-net with residual networks (U-ResNet) is employed for the task of dose prediction. The deep learning model is first trained from a database consisting of hundreds of historical treatment plans. The 3D dose distribution is then predicted with the input of the CT image and contours of regions of interest (ROIs). A total of 150 single-isocenter VMAT plans for multiple brain metastases are used for training and testing. The model performance is evaluated based on mean absolute error (MAE) and mean absolute differences of multiple dosimetric indexes (DIs), including (D max and D mean) for OARs, (D 98, D 95, D 50, and D 2) for PTVs, homogeneity index, and conformity index. The similarity between the predicted and clinically approved plan dose distribution is also evaluated.

Result

For 20 tested patients, the largest and smallest MAEs are 3.3% ± 3.6% and 1.3% ± 1.5%, respectively. The mean MAE for the 20 tested patients is 2.2% ± 0.7%. The mean absolute differences of D 98, D 95, D 50, and D2 for PTV60, PTV52, PTV50, and PTV40 are less than 2.5%, 3.0%, 2.0%, and 3.0%, respectively. The prediction accuracy of OARs for D max and D mean is within 3.2% and 1.2%, respectively. The average DSC ranges from 0.86 to 1 for all tested patients.

Conclusion

U-ResNet is viable to produce accurate dose distribution that is comparable to those of the clinically approved treatment plans. The predicted results can be used to improve current treatment planning design, plan quality, efficiency, etc.

Invertible and Variable Augmented Network for Pretreatment Patient-Specific Quality Assurance Dose Prediction

Pretreatment patient-specific quality assurance (prePSQA) is conducted to confirm the accuracy of the radiotherapy dose delivered. However, the process of prePSQA measurement is time consuming and exacerbates the workload for medical physicists. The purpose of this work is to propose a novel deep learning (DL) network to improve the accuracy and efficiency of prePSQA. A modified invertible and variable augmented network was developed to predict the three-dimensional (3D) measurement-guided dose (MDose) distribution of 300 cancer patients who underwent volumetric modulated arc therapy (VMAT) between 2018 and 2021, in which 240 cases were randomly selected for training, and 60 for testing. For simplicity, the present approach was termed as "IVPSQA." The input data include CT images, radiotherapy dose exported from the treatment planning system, and MDose distribution extracted from the verification system. Adam algorithm was used for first-order gradient-based optimization of stochastic objective functions. The IVPSQA model obtained high-quality 3D prePSQA dose distribution maps in head and neck, chest, and abdomen cases, and outperformed the existing U-Net-based prediction approaches in terms of dose difference maps and horizontal profiles comparison. Moreover, quantitative evaluation metrics including SSIM, MSE, and MAE demonstrated that the proposed approach achieved a good agreement with ground truth and yield promising gains over other advanced methods. This study presented the first work on predicting 3D prePSQA dose distribution by using the IVPSQA model. The proposed method could be taken as a clinical guidance tool and help medical physicists to reduce the measurement work of prePSQA.

Deep Learning-Based Prediction of Radiation Therapy Dose Distributions in Nasopharyngeal Carcinomas: A Preliminary Study Incorporating Multiple Features Including Images, Structures, and Dosimetry

Purpose: Intensity-modulated radiotherapy (IMRT) is currently the most important treatment method for nasopharyngeal carcinoma (NPC). This study aimed to enhance prediction accuracy by incorporating dose information into a deep convolutional neural network (CNN) using a multichannel input method. Methods: A target conformal plan (TCP) was created based on the maximum planning target volume (PTV). Input data included TCP dose distribution, images, target structures, and organ-at-risk (OAR) information. The role of target conformal plan dose (TCPD) was assessed by comparing the TCPD-CNN (with dose information) and NonTCPD-CNN models (without dose information) using statistical analyses with the ranked Wilcoxon test (P < .05 considered significant). Results: The TCPD-CNN model showed no statistical differences in predicted target indices, except for PTV60, where differences in the D98% indicator were < 0.5%. For OARs, there were no significant differences in predicted results, except for some small-volume or closely located OARs. On comparing TCPD-CNN and NonTCPD-CNN models, TCPD-CNN's dose-volume histograms closely resembled clinical plans with higher similarity index. Mean dose differences for target structures (predicted TCPD-CNN and NonTCPD-CNN results) were within 3% of the maximum prescription dose for both models. TCPD-CNN and NonTCPD-CNN outcomes were 67.9% and 54.2%, respectively. 3D gamma pass rates of the target structures and the entire body were higher in TCPD-CNN than in the NonTCPD-CNN models (P < .05). Additional evaluation on previously unseen volumetric modulated arc therapy plans revealed that average 3D gamma pass rates of the target structures were larger than 90%. Conclusions: This study presents a novel framework for dose distribution prediction using deep learning and multichannel input, specifically incorporating TCPD information, enhancing prediction accuracy for IMRT in NPC treatment.

Three-dimensional dose prediction based on deep convolutional neural networks for brain cancer in CyberKnife: accurate beam modelling of homogeneous tissue

Objectives

Accurate beam modelling is essential for dose calculation in stereotactic radiation therapy (SRT), such as CyberKnife treatment. However, the present deep learning methods only involve patient anatomical images and delineated masks for training. These studies generally focus on traditional intensity-modulated radiation therapy (RT) plans. Nevertheless, this paper aims to develop a deep CNN-based method for CyberKnife plan dose prediction about brain cancer patients. It utilized modelled beam information, target delineation, and patient anatomical information.

Methods

This study proposes a method that adds beam information to predict the dose distribution of CyberKnife in brain cases. A retrospective dataset of 88 brain and abdominal cancer patients treated with the Ray-tracing algorithm was performed. The datasets include patients' anatomical information (planning CT), binary masks for organs at risk (OARs) and targets, and clinical plans (containing beam information). The datasets were randomly split into 68, 6, and 14 brain cases for training, validation, and testing, respectively.

Results

Our proposed method performs well in SRT dose prediction. First, for the gamma passing rates in brain cancer cases, with the 2 mm/2% criteria, we got 96.7% ± 2.9% for the body, 98.3% ± 3.0% for the planning target volume, and 100.0% ± 0.0% for the OARs with small volumes referring to the clinical plan dose. Secondly, the model predictions matched the clinical plan's dose-volume histograms reasonably well for those cases. The differences in key metrics at the target area were generally below 1.0 Gy (approximately a 3% difference relative to the prescription dose).

Conclusions

The preliminary results for selected 14 brain cancer cases suggest that accurate 3-dimensional dose prediction for brain cancer in CyberKnife can be accomplished based on accurate beam modelling for homogeneous tumour tissue. More patients and other cancer sites are needed in a further study to validate the proposed method fully.

Advances in knowledge

With accurate beam modelling, the deep learning model can quickly generate the dose distribution for CyberKnife cases. This method accelerates the RT planning process, significantly improves its operational efficiency, and optimizes it.

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

PURPOSE

Ethos 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).

MATERIALS

General 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.

RESULTS

EG_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.

CONCLUSION

General 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.

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

BACKGROUND

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.

PURPOSE

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.

METHODS

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.

RESULTS

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 ± 0.46 Gy. The percentage dose differences in the predicted target metrics ofD 1 % ${D}_{1{\mathrm{\% }}}$ andD 98 % ${D}_{98{\mathrm{\% }}}$ 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 ± 1.66 Gy and -0.42 ± 2.07 Gy, 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.

CONCLUSIONS

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.

Patient selection for proton therapy using Normal Tissue Complication Probability with deep learning dose prediction for oropharyngeal cancer

BACKGROUND

In cancer care, determining the most beneficial treatment technique is a key decision affecting the patient's survival and quality of life. Patient selection for proton therapy (PT) over conventional radiotherapy (XT) currently entails comparing manually generated treatment plans, which requires time and expertise.

PURPOSE

We developed an automatic and fast tool, AI-PROTIPP (Artificial Intelligence Predictive Radiation Oncology Treatment Indication to Photons/Protons), that assesses quantitatively the benefits of each therapeutic option. Our method uses deep learning (DL) models to directly predict the dose distributions for a given patient for both XT and PT. By using models that estimate the Normal Tissue Complication Probability (NTCP), namely the likelihood of side effects to occur for a specific patient, AI-PROTIPP can propose a treatment selection quickly and automatically.

METHODS

A database of 60 patients presenting oropharyngeal cancer, obtained from the Cliniques Universitaires Saint Luc in Belgium, was used in this study. For every patient, a PT plan and an XT plan were generated. The dose distributions were used to train the two dose DL prediction models (one for each modality). The model is based on U-Net architecture, a type of convolutional neural network currently considered as the state of the art for dose prediction models. A NTCP protocol used in the Dutch model-based approach, including grades II and III xerostomia and grades II and III dysphagia, was later applied in order to perform automatic treatment selection for each patient. The networks were trained using a nested cross-validation approach with 11-folds. We set aside three patients in an outer set and each fold consists of 47 patients in training, five in validation and five for testing. This method allowed us to assess our method on 55 patients (five patients per test times the number of folds).

RESULTS

The treatment selection based on the DL-predicted doses reached an accuracy of 87.4% for the threshold parameters set by the Health Council of the Netherlands. The selected treatment is directly linked with these threshold parameters as they express the minimal gain brought by the PT treatment for a patient to be indicated to PT. To validate the performance of AI-PROTIPP in other conditions, we modulated these thresholds, and the accuracy was above 81% for all the considered cases. The difference in average cumulative NTCP per patient of predicted and clinical dose distributions is very similar (less than 1% difference).

CONCLUSIONS

AI-PROTIPP shows that using DL dose prediction in combination with NTCP models to select PT for patients is feasible and can help to save time by avoiding the generation of treatment plans only used for the comparison. Moreover, DL models are transferable, allowing, in the future, experience to be shared with centers that would not have PT planning expertise.

A deep-learning-based dose verification tool utilizing fluence maps for a cobalt-60 compensator-based intensity-modulated radiation therapy system

Background and purpose

A novel cobalt-60 compensator-based intensity-modulated radiation therapy (IMRT) system was developed for a resource-limited environment but lacked an efficient dose verification algorithm. The aim of this study was to develop a deep-learning-based dose verification algorithm for accurate and rapid dose predictions.

Materials and methods

A deep-learning network was employed to predict the doses from static fields related to beam commissioning. Inputs were a cube-shaped phantom, a beam binary mask, and an intersecting volume of the phantom and beam binary mask, while output was a 3-dimensional (3D) dose. The same network was extended to predict patient-specific doses for head and neck cancers using two different approaches. A field-based method predicted doses for each field and combined all calculated doses into a plan, while the plan-based method combined all nine fluences into a plan to predict doses. Inputs included patient computed tomography (CT) scans, binary beam masks, and fluence maps truncated to the patient's CT in 3D.

Results

For static fields, predictions agreed well with ground truths with average deviations of less than 0.5% for percent depth doses and profiles. Even though the field-based method showed excellent prediction performance for each field, the plan-based method showed better agreement between clinical and predicted dose distributions. The distributed dose deviations for all planned target volumes and organs at risk were within 1.3 Gy. The calculation speed for each case was within two seconds.

Conclusions

A deep-learning-based dose verification tool can accurately and rapidly predict doses for a novel cobalt-60 compensator-based IMRT system.

Evaluating machine learning enhanced intelligent-optimization-engine (IOE) performance for ethos head-and-neck (HN) plan generation

PURPOSE

Varian Ethos utilizes novel intelligent-optimization-engine (IOE) designed to automate the planning. However, this introduced a black box approach to plan optimization and challenge for planners to improve plan quality. This study aims to evaluate machine-learning-guided initial reference plan generation approaches for head & neck (H&N) adaptive radiotherapy (ART).

METHODS

Twenty previously treated patients treated on C-arm/Ring-mounted were retroactively re-planned in the Ethos planning system using a fixed 18-beam intensity-modulated radiotherapy (IMRT) template. Clinical goals for IOE input were generated using (1) in-house deep-learning 3D-dose predictor (AI-Guided) (2) commercial knowledge-based planning (KBP) model with universal RTOG-based population criteria (KBP-RTOG) and (3) an RTOG-based constraint template only (RTOG) for in-depth analysis of IOE sensitivity. Similar training data was utilized for both models. Plans were optimized until their respective criteria were achieved or DVH-estimation band was satisfied. Plans were normalized such that the highest PTV dose level received 95% coverage. Target coverage, high-impact organs-at-risk (OAR) and plan deliverability was assessed in comparison to clinical (benchmark) plans. Statistical significance was evaluated using a paired two-tailed student t-test.

RESULTS

AI-guided plans were superior to both KBP-RTOG and RTOG-only plans with respect to clinical benchmark cases. Overall, OAR doses were comparable or improved with AI-guided plans versus benchmark, while they increased with KBP-RTOG and RTOG plans. However, all plans generally satisfied the RTOG criteria. Heterogeneity Index (HI) was on average <1.07 for all plans. Average modulation factor was 12.2 ± 1.9 (p = n.s), 13.1 ± 1.4 (p = <0.001), 11.5 ± 1.3 (p = n.s.) and 12.2 ± 1.9 for KBP-RTOG, AI-Guided, RTOG and benchmark plans, respectively.

CONCLUSION

AI-guided plans were the highest quality. Both KBP-enabled and RTOG-only plans are feasible approaches as clinics adopt ART workflows. Similar to constrained optimization, the IOE is sensitive to clinical input goals and we recommend comparable input to an institution's planning directive dosimetric criteria.

A convolutional neural network model for EPID-based non-transit dosimetry

PURPOSE

To develop an alternative computational approach for EPID-based non-transit dosimetry using a convolutional neural network model.

METHOD

A U-net followed by a non-trainable layer named True Dose Modulation recovering the spatialized information was developed. The model was trained on 186 Intensity-Modulated Radiation Therapy Step & Shot beams from 36 treatment plans of different tumor locations to convert grayscale portal images into planar absolute dose distributions. Input data were acquired from an amorphous-Silicon Electronic Portal Image Device and a 6 MV X-ray beam. Ground truths were computed from a conventional kernel-based dose algorithm. The model was trained by a two-step learning process and validated through a five-fold cross-validation procedure with sets of training and validation of 80% and 20%, respectively. A study regarding the dependance of the amount of training data was conducted. The performance of the model was evaluated from a quantitative analysis based the ϒ-index, absolute and relative errors computed between the inferred dose distributions and ground truths for six square and 29 clinical beams from seven treatment plans. These results were also compared to those of an existing portal image-to-dose conversion algorithm.

RESULTS

For the clinical beams, averages of ϒ-index and ϒ-passing rate (2%-2mm > 10% D_max ) of 0.24 (±0.04) and 99.29 (±0.70)% were obtained. For the same metrics and criteria, averages of 0.31 (±0.16) and 98.83 (±2.40)% were obtained with the six square beams. Overall, the developed model performed better than the existing analytical method. The study also showed that sufficient model accuracy can be achieved with the amount of training samples used.

CONCLUSION

A deep learning-based model was developed to convert portal images into absolute dose distributions. The accuracy obtained shows that this method has great potential for EPID-based non-transit dosimetry.

Artificial Intelligence Guided Physician Directive Improves Head and Neck Planning Quality and Practice Uniformity: A Prospective Study

•AI dose predictor was fully integrated with treatment planning system and used as a physicain decision support tool to improve uniformity of practice.•Model was trained based on our standard of practice, but implemented at the time of expansion with 3 new physicians join the practice.•Phase 1 retrospective evaluation demonstrated the non-uniform practice among 3 MDs and only 52.9% frequency planner can achieve physicians' directives.•Significant improvement in practice uniformity of practice was observed after utilizing AI as DST and 80.4% frequency clinical plan can achieve AI-guided physician directives.

Deep learning-based prediction of deliverable adaptive plans for MR-guided adaptive radiotherapy: A feasibility study

Purpose

Fast and automated plan generation is desirable in radiation therapy (RT), in particular, for MR-guided online adaptive RT (MRgOART) or real-time (intrafractional) adaptive RT (MRgRART), to reduce replanning time. The purpose of this study is to investigate the feasibility of using deep learning to quickly predict deliverable adaptive plans based on a target dose distribution for MRgOART/MRgRART.

Methods

A conditional generative adversarial network (cGAN) was trained to predict the MLC leaf sequence corresponding to a target dose distribution based on reference plan created prior to MRgOART using a 1.5T MR-Linac. The training dataset included 50 ground truth dose distributions and corresponding beam parameters (aperture shapes and weights) created during MRgOART for 10 pancreatic cancer patients (each with five fractions). The model input was the dose distribution from each individual beam and the output was the predicted corresponding field segments with specific shape and weight. Patient-based leave-one-out-cross-validation was employed and for each model trained, four (44 training beams) out of five fractionated plans of the left-out patient were set aside for testing purposes. We deliberately kept a single fractionated plan in the training dataset so that the model could learn to replan the patient based on a prior plan. The model performance was evaluated by calculating the gamma passing rate of the ground truth dose vs. the dose from the predicted adaptive plan and calculating max and mean dose metrics.

Results

The average gamma passing rate (95%, 3mm/3%) among 10 test cases was 88%. In general, we observed 95% of the prescription dose to PTV achieved with an average 7.6% increase of max and mean dose, respectively, to OARs for predicted replans. Complete adaptive plans were predicted in ≤20 s using a GTX 1660TI GPU.

Conclusion

We have proposed and demonstrated a deep learning method to generate adaptive plans automatically and rapidly for MRgOART. With further developments using large datasets and the inclusion of patient contours, the method may be implemented to accelerate MRgOART process or even to facilitate MRgRART.

Translating Data Science Results into Precision Oncology Decisions: A Mini Review

While reviewing and discussing the potential of data science in oncology, we emphasize medical imaging and radiomics as the leading contextual frameworks to measure the impacts of Artificial Intelligence (AI) and Machine Learning (ML) developments. We envision some domains and research directions in which radiomics should become more significant in view of current barriers and limitations.

Automation of pencil beam scanning proton treatment planning for intracranial tumours

PURPOSE

To evaluate the feasibility of comprehensive automation of an intra-cranial proton treatment planning.

MATERIALS AND METHODS

Class solution (CS) beam configuration selection allows the user to identify predefined beam configuration based on target localization; automatic CS (aCS) will then explore all the possible CS beam geometries. Ten patients, already used for the evaluation of the automatic selection of the beam configuration, have been also employed to training an algorithm based on the computation of a benchmark dose exploit automatic general planning solution (GPS) optimization with a wish list approach for the planning optimization. An independent cohort of ten patients has been then used for the evaluation step between the clinical and the GPS plan in terms of dosimetric quality of plans and the time needed to generate a plan.

RESULTS

The definition of a beam configuration requires on average 22 min (range 9-29 min). The average time for GPS plan generation is 18 min (range 7-26 min). Median dose differences (GPS-Manual) for each OAR constraints are: brainstem -1.60 Gy, left cochlea -1.22 Gy, right cochlea -1.42 Gy, left eye 0.55 Gy, right eye -2.33 Gy, optic chiasm -1.87 Gy, left optic nerve -4.45 Gy, right optic nerve -2.48 Gy and optic tract -0.31 Gy. Dosimetric CS and aCS plan evaluation shows a slightly worsening of the OARs values except for the optic tract and optic chiasm for both CS and aCS, where better results have been observed.

CONCLUSION

This study has shown the feasibility and implementation of the automatic planning system for intracranial tumors. The method developed in this work is ready to be implemented in a clinical workflow.

Advances in Automated Treatment Planning

Treatment planning in radiation therapy has progressed enormously over the past several decades. Such advancements came in the form of innovative hardware and algorithms, giving rise to modalities such as intensity-modulated radiation therapy and volume modulated arc therapy, greatly improving patient outcome and quality of life. While these developments have improved the overall plan quality, they have also given rise to higher treatment planning complexity. This has resulted in increased treatment planning time and higher variability in the final approved plan quality. Radiation oncology, as an already technologically advanced field, has much research and implementation involving the use of AI. The field has begun to show the efficacy of using such technologies in many of its sub-areas, such as in diagnosis, imaging, segmentation, treatment planning, quality assurance, treatment delivery, and follow-up. Some AI technologies have already been clinically implemented by commercial systems. In this article, we will provide an overview to methods involved with treatment planning in radiation therapy. In particular, we will review the recent research and literature related to automation of the treatment planning process, leading to potentially higher efficiency and higher quality plans. We will then present the current and future challenges, as well as some future perspectives.

A novel multichannel deep learning model for fast denoising of Monte Carlo dose calculations: preclinical applications

Objective. In preclinical radiotherapy with kilovolt (kV) x-ray beams, accurate treatment planning is needed to improve the translation potential to clinical trials. Monte Carlo based radiation transport simulations are the gold standard to calculate the absorbed dose distribution in external beam radiotherapy. However, these simulations are notorious for their long computation time, causing a bottleneck in the workflow. Previous studies have used deep learning models to speed up these simulations for clinical megavolt (MV) beams. For kV beams, dose distributions are more affected by tissue type than for MV beams, leading to steep dose gradients. This study aims to speed up preclinical kV dose simulations by proposing a novel deep learning pipeline. Approach. A deep learning model is proposed that denoises low precision (∼106 simulated particles) dose distributions to produce high precision (109 simulated particles) dose distributions. To effectively denoise the steep dose gradients in preclinical kV dose distributions, the model uses the novel approach to use the low precision Monte Carlo dose calculation as well as the Monte Carlo uncertainty (MCU) map and the mass density map as additional input channels. The model was trained on a large synthetic dataset and tested on a real dataset with a different data distribution. To keep model inference time to a minimum, a novel method for inference optimization was developed as well. Main results. The proposed model provides dose distributions which achieve a median gamma pass rate (3%/0.3 mm) of 98% with a lower bound of 95% when compared to the high precision Monte Carlo dose distributions from the test set, which represents a different dataset distribution than the training set. Using the proposed model together with the novel inference optimization method, the total computation time was reduced from approximately 45 min to less than six seconds on average. Significance. This study presents the first model that can denoise preclinical kV instead of clinical MV Monte Carlo dose distributions. This was achieved by using the MCU and mass density maps as additional model inputs. Additionally, this study shows that training such a model on a synthetic dataset is not only a viable option, but even increases the generalization of the model compared to training on real data due to the sheer size and variety of the synthetic dataset. The application of this model will enable speeding up treatment plan optimization in the preclinical workflow.

Artificial intelligence in radiotherapy

Radiotherapy is a discipline closely integrated with computer science. Artificial intelligence (AI) has developed rapidly over the past few years. With the explosive growth of medical big data, AI promises to revolutionize the field of radiotherapy through highly automated workflow, enhanced quality assurance, improved regional balances of expert experiences, and individualized treatment guided by multi-omics. In addition to independent researchers, the increasing number of large databases, biobanks, and open challenges significantly facilitated AI studies on radiation oncology. This article reviews the latest research, clinical applications, and challenges of AI in each part of radiotherapy including image processing, contouring, planning, quality assurance, motion management, and outcome prediction. By summarizing cutting-edge findings and challenges, we aim to inspire researchers to explore more future possibilities and accelerate the arrival of AI radiotherapy.

A Survey on Deep Learning for Precision Oncology

Precision oncology, which ensures optimized cancer treatment tailored to the unique biology of a patient’s disease, has rapidly developed and is of great clinical importance. Deep learning has become the main method for precision oncology. This paper summarizes the recent deep-learning approaches relevant to precision oncology and reviews over 150 articles within the last six years. First, we survey the deep-learning approaches categorized by various precision oncology tasks, including the estimation of dose distribution for treatment planning, survival analysis and risk estimation after treatment, prediction of treatment response, and patient selection for treatment planning. Secondly, we provide an overview of the studies per anatomical area, including the brain, bladder, breast, bone, cervix, esophagus, gastric, head and neck, kidneys, liver, lung, pancreas, pelvis, prostate, and rectum. Finally, we highlight the challenges and discuss potential solutions for future research directions.

Towards a safe and efficient clinical implementation of machine learning in radiation oncology by exploring model interpretability, explainability and data-model dependency

The interest in machine learning (ML) has grown tremendously in recent years, partly due to the performance leap that occurred with new techniques of deep learning, convolutional neural networks for images, increased computational power, and wider availability of large datasets. Most fields of medicine follow that popular trend and, notably, radiation oncology is one of those that are at the forefront, with already a long tradition in using digital images and fully computerized workflows. ML models are driven by data, and in contrast with many statistical or physical models, they can be very large and complex, with countless generic parameters. This inevitably raises two questions, namely, the tight dependence between the models and the datasets that feed them, and the interpretability of the models, which scales with its complexity. Any problems in the data used to train the model will be later reflected in their performance. This, together with the low interpretability of ML models, makes their implementation into the clinical workflow particularly difficult. Building tools for risk assessment and quality assurance of ML models must involve then two main points: interpretability and data-model dependency. After a joint introduction of both radiation oncology and ML, this paper reviews the main risks and current solutions when applying the latter to workflows in the former. Risks associated with data and models, as well as their interaction, are detailed. Next, the core concepts of interpretability, explainability, and data-model dependency are formally defined and illustrated with examples. Afterwards, a broad discussion goes through key applications of ML in workflows of radiation oncology as well as vendors' perspectives for the clinical implementation of ML.

Attention-aware 3D U-Net convolutional neural network for knowledge-based planning 3D dose distribution prediction of head-and-neck cancer

PURPOSE

Deep learning-based knowledge-based planning (KBP) methods have been introduced for radiotherapy dose distribution prediction to reduce the planning time and maintain consistent high-quality plans. This paper presents a novel KBP model using an attention-gating mechanism and a three-dimensional (3D) U-Net for intensity-modulated radiation therapy (IMRT) 3D dose distribution prediction in head-and-neck cancer.

METHODS

A total of 340 head-and-neck cancer plans, representing the OpenKBP-2020 AAPM Grand Challenge data set, were used in this study. All patients were treated with the IMRT technique and a dose prescription of 70 Gy. The data set was randomly divided into 64%/16%/20% as training/validation/testing cohorts. An attention-gated 3D U-Net architecture model was developed to predict full 3D dose distribution. The developed model was trained using the mean-squared error loss function, Adam optimization algorithm, a learning rate of 0.001, 120 epochs, and batch size of 4. In addition, a baseline U-Net model was also similarly trained for comparison. The model performance was evaluated on the testing data set by comparing the generated dose distributions against the ground-truth dose distributions using dose statistics and clinical dosimetric indices. Its performance was also compared to the baseline model and the reported results of other deep learning-based dose prediction models.

RESULTS

The proposed attention-gated 3D U-Net model showed high capability in accurately predicting 3D dose distributions that closely replicated the ground-truth dose distributions of 68 plans in the test set. The average value of the mean absolute dose error was 2.972 ± 1.220 Gy (vs. 2.920 ± 1.476 Gy for a baseline U-Net) in the brainstem, 4.243 ± 1.791 Gy (vs. 4.530 ± 2.295 Gy for a baseline U-Net) in the left parotid, 4.622 ± 1.975 Gy (vs. 4.223 ± 1.816 Gy for a baseline U-Net) in the right parotid, 3.346 ± 1.198 Gy (vs. 2.958 ± 0.888 Gy for a baseline U-Net) in the spinal cord, 6.582 ± 3.748 Gy (vs. 5.114 ± 2.098 Gy for a baseline U-Net) in the esophagus, 4.756 ± 1.560 Gy (vs. 4.992 ± 2.030 Gy for a baseline U-Net) in the mandible, 4.501 ± 1.784 Gy (vs. 4.925 ± 2.347 Gy for a baseline U-Net) in the larynx, 2.494 ± 0.953 Gy (vs. 2.648 ± 1.247 Gy for a baseline U-Net) in the PTV_70, and 2.432 ± 2.272 Gy (vs. 2.811 ± 2.896 Gy for a baseline U-Net) in the body contour. The average difference in predicting the D₉₉ value for the targets (PTV_70, PTV_63, and PTV_56) was 2.50 ± 1.77 Gy. For the organs at risk, the average difference in predicting the D m a x ${D_{max}}$ (brainstem, spinal cord, and mandible) and D m e a n ${D_{mean}}$ (left parotid, right parotid, esophagus, and larynx) values was 1.43 ± 1.01 and 2.44 ± 1.73 Gy, respectively. The average value of the homogeneity index was 7.99 ± 1.45 for the predicted plans versus 5.74 ± 2.95 for the ground-truth plans, whereas the average value of the conformity index was 0.63 ± 0.17 for the predicted plans versus 0.89 ± 0.19 for the ground-truth plans. The proposed model needs less than 5 s to predict a full 3D dose distribution of 64 × 64 × 64 voxels for a new patient that is sufficient for real-time applications.

CONCLUSIONS

The attention-gated 3D U-Net model demonstrated a capability in predicting accurate 3D dose distributions for head-and-neck IMRT plans with consistent quality. The prediction performance of the proposed model was overall superior to a baseline standard U-Net model, and it was also competitive to the performance of the best state-of-the-art dose prediction method reported in the literature. The proposed model could be used to obtain dose distributions for decision-making before planning, quality assurance of planning, and guiding-automated planning for improved plan consistency, quality, and planning efficiency.

Millisecond speed deep learning based proton dose calculation with Monte Carlo accuracy

Objective.Next generation online and real-time adaptive radiotherapy workflows require precise particle transport simulations in sub-second times, which is unfeasible with current analytical pencil beam algorithms (PBA) or Monte Carlo (MC) methods. We present a deep learning based millisecond speed dose calculation algorithm (DoTA) accurately predicting the dose deposited by mono-energetic proton pencil beams for arbitrary energies and patient geometries.Approach.Given the forward-scattering nature of protons, we frame 3D particle transport as modeling a sequence of 2D geometries in the beam's eye view. DoTA combines convolutional neural networks extracting spatial features (e.g. tissue and density contrasts) with a transformer self-attention backbone that routes information between the sequence of geometry slices and a vector representing the beam's energy, and is trained to predict low noise MC simulations of proton beamlets using 80 000 different head and neck, lung, and prostate geometries.Main results.Predicting beamlet doses in 5 ± 4.9 ms with a very high gamma pass rate of 99.37 ± 1.17% (1%, 3 mm) compared to the ground truth MC calculations, DoTA significantly improves upon analytical pencil beam algorithms both in precision and speed. Offering MC accuracy 100 times faster than PBAs for pencil beams, our model calculates full treatment plan doses in 10-15 s depending on the number of beamlets (800-2200 in our plans), achieving a 99.70 ± 0.14% (2%, 2 mm) gamma pass rate across 9 test patients.Significance.Outperforming all previous analytical pencil beam and deep learning based approaches, DoTA represents a new state of the art in data-driven dose calculation and can directly compete with the speed of even commercial GPU MC approaches. Providing the sub-second speed required for adaptive treatments, straightforward implementations could offer similar benefits to other steps of the radiotherapy workflow or other modalities such as helium or carbon treatments.

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

INTRODUCTION

Intensity 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/METHODS

Robustly 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.

RESULTS

An 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.

CONCLUSION

Robust 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.

Accuracy Improvement Method Based on Characteristic Database Classification for IMRT Dose Prediction in Cervical Cancer: Scientifically Training Data Selection

Purpose

Consistent training and testing datasets can lead to good performance for deep learning (DL) models. However, a large high-quality training dataset for unusual clinical scenarios is usually not easy to collect. The work aims to find optimal training data collection strategies for DL-based dose prediction models.

Materials and Methods

A total of 325 clinically approved cervical IMRT plans were utilized. We designed comparison experiments to investigate the impact of (1) beam angles, (2) the number of beams, and (3) patient position for DL dose prediction models. In addition, a novel geometry-based beam mask generation method was proposed to provide beam setting information in the model training process. What is more, we proposed a new training strategy named “full-database pre-trained strategy”.

Results

The model trained with a homogeneous dataset with the same beam settings achieved the best performance [mean prediction errors of planning target volume (PTV), bladder, and rectum: 0.29 ± 0.15%, 3.1 ± 2.55%, and 3.15 ± 1.69%] compared with that trained with large mixed beam setting plans (mean errors of PTV, bladder, and rectum: 0.8 ± 0.14%, 5.03 ± 2.2%, and 4.45 ± 1.4%). A homogeneous dataset is more accessible to train an accurate dose prediction model (mean errors of PTV, bladder and rectum: 2.2 ± 0.15%, 5 ± 2.1%, and 3.23 ± 1.53%) than a non-homogeneous one (mean errors of PTV, bladder and rectum: 2.55 ± 0.12%, 6.33 ± 2.46%, and 4.76 ± 2.91%) without other processing approaches. The added beam mask can constantly improve the model performance, especially for datasets with different beam settings (mean errors of PTV, bladder, and rectum improved from 0.8 ± 0.14%, 5.03 ± 2.2%, and 4.45 ± 1.4% to 0.29 ± 0.15%, 3.1 ± 2.55%, and 3.15 ± 1.69%).

Conclusions

A consistent dataset is recommended to form a patient-specific IMRT dose prediction model. When a consistent dataset is not accessible to collect, a large dataset with different beam angles and a training model with beam information can also get a relatively good model. The full-database pre-trained strategies can rapidly form an accuracy model from a pre-trained model. The proposed beam mask can effectively improve the model performance. Our study may be helpful for further dose prediction studies in terms of training strategies or database establishment.

Robust automated radiation therapy treatment planning using scenario-specific dose prediction and robust dose mimicking

Purpose

We present a framework for robust automated treatment planning using machine learning, comprising scenario‐specific dose prediction and robust dose mimicking.

Methods

The scenario dose prediction pipeline is divided into the prediction of nominal dose from input image and the prediction of scenario dose from nominal dose, each using a deep learning model with U‐net architecture. By using a specially developed dose–volume histogram–based loss function, the predicted scenario doses are ensured sufficient target coverage despite the possibility of the training data being non‐robust. Deliverable plans may then be created by solving a robust dose mimicking problem with the predictions as scenario‐specific reference doses.

Results

Numerical experiments are performed using a data set of 52 intensity‐modulated proton therapy plans for prostate patients. We show that the predicted scenario doses resemble their respective ground truth well, in particular while having target coverage comparable to that of the nominal scenario. The deliverable plans produced by the subsequent robust dose mimicking were showed to be robust against the same scenario set considered for prediction.

Conclusions

We demonstrate the feasibility and merits of the proposed methodology for incorporating robustness into automated treatment planning algorithms.

The feasibility study on the generalization of deep learning dose prediction model for volumetric modulated arc therapy of cervical cancer

Purpose

To develop a 3D‐Unet dose prediction model to predict the three‐dimensional dose distribution of volumetric modulated arc therapy (VMAT) for cervical cancer and test the dose prediction performance of the model in endometrial cancer to explore the feasibility of model generalization.

Methods

One hundred and seventeen cases of cervical cancer and 20 cases of endometrial cancer treated with VMAT were used for the model training, validation, and test. The prescribed dose was 50.4 Gy in 28 fractions. Eight independent channels of contoured structures were input to the model, and the dose distribution was used as the output of the model. The 3D‐Unet prediction model was trained and validated on the training set (n = 86) and validation set (n = 11), respectively. Then the model was tested on the test set (n = 20) of cervical cancer and endometrial cancer, respectively. The results between clinical dose distribution and predicted dose distribution were compared in the following aspects: (a) the mean absolute error (MAE) within the body, (b) the Dice similarity coefficients (DSCs) under different isodose volumes, (c) the dosimetric indexes including the mean dose (D_mean), the received dose of 2 cm³ (D_2cc), the percentage volume of receiving 40 Gy dose of organs‐at‐risk (V₄₀), planning target volume (PTV) D_98%, and homogeneity index (HI), (d) dose–volume histograms (DVHs).

Results

The model can accurately predict the dose distribution of the VMAT plan for cervical cancer and endometrial cancer. The overall average MAE and maximum MAE for cervical cancer were 2.43 ± 3.17% and 3.16 ± 4.01% of the prescribed dose, respectively, and for endometrial cancer were 2.70 ± 3.54% and 3.85 ± 3.11%. The average DSCs under different isodose volumes is above 0.9. The predicted dosimetric indexes and DVHs are equivalent to the clinical dose for both cervical cancer and endometrial cancer, and there is no statistically significant difference.

Conclusion

A 3D‐Unet dose prediction model was developed for VMAT of cervical cancer, which can predict the dose distribution accurately for cervical cancer. The model can also be generalized for endometrial cancer with good performance.

Site-agnostic 3D dose distribution prediction with deep learning neural networks

PURPOSE

Typically, the current dose prediction models are limited to small amounts of data and require retraining for a specific site, often leading to suboptimal performance. We propose a site-agnostic, three-dimensional dose distribution prediction model using deep learning that can leverage data from any treatment site, thus increasing the total data available to train the model. Applying our proposed model to a new target treatment site requires only a brief fine-tuning of the model to the new data and involves no modifications to the model input channels or its parameters. Thus, it can be efficiently adapted to a different treatment site, even with a small training dataset.

METHODS

This study uses two separate datasets/treatment sites: data from patients with prostate cancer treated with intensity-modulated radiation therapy (source data), and data from patients with head-and-neck cancer treated with volumetric-modulated arc therapy (target data). We first developed a source model with 3D UNet architecture, trained from random initial weights on the source data. We evaluated the performance of this model on the source data. We then studied the generalizability of the model to the new target dataset via transfer learning. To do this, we built three more models, all with the same 3D UNet architecture: target model, adapted model, and combined model. The source and target models were trained on the source and target data from random initial weights, respectively. The adapted model fine-tuned the source model to the target domain by using the target data. Finally, the combined model was trained from random initial weights on a combined data pool consisting of both target and source datasets. We tested all four models on the target dataset and evaluated quantitative dose-volume histogram metrics for the planning target volume (PTV) and organs at risk (OARs).

RESULTS

When tested on the source treatment site, the source model accurately predicted the dose distributions with average (mean, max) absolute dose errors of (0.32%±0.14, 2.37%±0.93) (PTV) relative to the prescription dose, and highest mean dose error of 1.68%±0.76, and highest max dose error of 5.47%± 3.31 for femoral head right. The error in PTV dose coverage prediction is 3.21%±1.51 for D₉₈ , 3.04%±1.69 for D₉₅ , and 1.83%±1.01 for D₀₂ . Averaging across all OARs, the source model predicted the OAR mean dose within 1.38% and the OAR max dose within 3.64%. For the target treatment site, the target model average (mean, max) absolute dose errors relative to the prescription dose for the PTV were (1.08%±0.95, 2.90%±1.35). Left cochlea had the highest mean and max dose errors of 5.37%±5.82 and 8.33%±8.88, respectively. The errors in PTV dose coverage prediction for D₉₈ and D₉₅ were 2.88%±1.59 and 2.55%±1.28, respectively. The target model can predict the OAR mean dose within 2.43% and the OAR max dose within 4.33% on average across all OARs.

CONCLUSION

We developed a site-agnostic model for three-dimensional dose prediction and tested its adaptability to a new target treatment site via transfer learning. Our proposed model can make accurate predictions with limited training data.

Volumetric dose extension for isodose tuning

PURPOSE

To develop a method that can extend dose from two isodose surfaces (isosurfaces) to the entire patient volume, and to demonstrate its application in radiotherapy plan isodose tuning.

METHODS

We hypothesized that volumetric dose distribution can be extended from two isosurfaces-the 100% isosurface and a reference isosurface--with the distances to these two surfaces (L_ref ) as extension variables. The extension function is modeled by a three-dimensional lookup table (LUT), where voxel dose values from clinical plans are binned by three indexes: L₁₀₀ , L_ref and D_ref (reference dose level). The mean and standard deviation of voxel doses in each bin are calculated and stored in LUT. Volumetric dose extension is performed voxelwisely by indexing the LUT with the L₁₀₀ , L_ref and D_ref of each query voxel. The mean dose stored in the corresponding bin is filled into the query voxel as extended dose, and the standard deviation be filled voxelwisely as the uncertainty of extension result. We applied dose extension in isodose tuning, which aims to tune volumetric dose distribution by isosurface dragging. We adopted extended dose as an approximate dose estimation, and combined it with dose correction strategy to achieve accurate dose tuning.

RESULTS

We collected 32 post-operative prostate volumetric modulated arc therapy (VMAT) cases and built the LUT and its associated uncertainties from the doses of 27 cases. The dose extension method was tested on five cases, whose dose distributions were defined as ground truth (GT). We extended the doses from 100% and 50% GT isosurfaces to the entire volume, and evaluated the accuracy of extended doses. The 5 mm/5% gamma passing rate (GPR) of extended doses are 92.0%. The mean error is 4.5%, which is consistent to the uncertainty estimated by LUT. The dose difference in 90.5% of voxels is within two sigma and 97.5% in three sigma. The calculation time is less than two seconds. To simulate plan isodose tuning, we optimized a dose with less sparing on rectum (than GT dose) and defined it as a "base dose"-the dose awaiting isosurface dragging. In front-end, the simulated isodose tuning is conducted as such that the base dose was given to plan tuner, and its 50% isosurface would be dragged to the desired position (position of 50% isosurface in GT dose). In back-end, the output of isodose tuning is obtained by 1) extending dose from the desired isosurfaces and viewed the extended dose as an approximate dose, 2) obtaining a correction map from the base dose, and 3) applying the correction map to the extended dose. The accuracy of output-extended dose with correction-was 97.2% in GPR (3 mm/3%) and less than 1% in mean dose difference. The total calculation time is less than two seconds, which allows for interactive isodose tuning.

CONCLUSIONS

We developed a dose extension method that generates volumetric dose distribution from two surfaces. The application of dose extension is in interactive isodose tuning. The distance-based LUT fashion and correction strategy guarantee the computation efficiency and accuracy in isodose tuning. This article is protected by copyright. All rights reserved.

Artificial Intelligence in Radiation Therapy

Artificial intelligence (AI) has great potential to transform the clinical workflow of radiotherapy. Since the introduction of deep neural networks, many AI-based methods have been proposed to address challenges in different aspects of radiotherapy. Commercial vendors have started to release AI-based tools that can be readily integrated to the established clinical workflow. To show the recent progress in AI-aided radiotherapy, we have reviewed AI-based studies in five major aspects of radiotherapy including image reconstruction, image registration, image segmentation, image synthesis, and automatic treatment planning. In each section, we summarized and categorized the recently published methods, followed by a discussion of the challenges, concerns, and future development. Given the rapid development of AI-aided radiotherapy, the efficiency and effectiveness of radiotherapy in the future could be substantially improved through intelligent automation of various aspects of radiotherapy.

A hybrid optimization strategy for deliverable Intensity-modulated radiotherapy plan generation using deep learning-based dose prediction

PURPOSE

To propose a clinically feasible automatic planning solution for external beam intensity-modulated radiotherapy (IMRT) including dose prediction via a deep learning and voxel-based optimization strategy.

MATERIALS AND METHODS

The dose distribution of patients was predicted using a U-Net-based deep learning network based on the patient's anatomy information. One hundred seventeen patients with nasopharyngeal cancer (NPC) and 200 patients with rectal cancer were enrolled in this study. For NPC cases, 94 cases were included in the training dataset, 13 in the validation dataset and 10 in the testing dataset. For rectal cancer cases, 172 cases were included in the training set, 18 in the validation set and 10 in the testing set. A voxel-based optimization strategy, "Voxel", was proposed to achieve treatment planning optimization by dividing body voxels into two parts: inside planning target volumes (PTVs) and outside PTVs. Fixed dose-volume objectives were attached to the total objective function to realize individualized planning intended as the "hybrid" optimizing strategy. Automatically generated plans were compared with clinically approved plans to evaluate clinical gains, according to dosimetric indices and dose-volume histograms (DVHs).

RESULTS

Similarities were found between the DVH of the predicted dose and clinical plan, although significant differences were found in some organs at risk (OARs). Better organ sparing and suboptimal PTV coverage were shown using the voxel strategy; however, the deviation in homogeneity indices (HI) and conformity indices (CI) of the PTV between automatically generated plans and manual plans were reduced by the hybrid strategy ((Manual plans)/(Voxel plans)/(Hybrid plans): HI of PTV70 (1.06/1.12/1.02), CI of PTV70 (0.79/0.58/0.76)). The optimization time for each patient was within 1 minute and included fluence map optimization, leaf sequencing and control point optimization. All the generated plans (voxel & hybrid strategy) could be delivered on uRT-linac 506c (United Imaging Health care, Shanghai, China).

CONCLUSION

Deliverable plans can be generated by incorporating a voxel-based optimization strategy into a commercial TPS. The hybrid optimization method shows the benefit and clinical feasibility in generating clinically acceptable plans. This article is protected by copyright. All rights reserved.