Models for predicting objective function weights in prostate cancer IMRT

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

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

Artificial general intelligence for radiation oncology

The emergence of artificial general intelligence (AGI) is transforming radiation oncology. As prominent vanguards of AGI, large language models (LLMs) such as GPT-4 and PaLM 2 can process extensive texts and large vision models (LVMs) such as the Segment Anything Model (SAM) can process extensive imaging data to enhance the efficiency and precision of radiation therapy. This paper explores full-spectrum applications of AGI across radiation oncology including initial consultation, simulation, treatment planning, treatment delivery, treatment verification, and patient follow-up. The fusion of vision data with LLMs also creates powerful multimodal models that elucidate nuanced clinical patterns. Together, AGI promises to catalyze a shift towards data-driven, personalized radiation therapy. However, these models should complement human expertise and care. This paper provides an overview of how AGI can transform radiation oncology to elevate the standard of patient care in radiation oncology, with the key insight being AGI's ability to exploit multimodal clinical data at scale.

Implementation and evaluation of an intelligent automatic treatment planning robot for prostate cancer stereotactic body radiation therapy

PURPOSE

We previously developed a virtual treatment planner (VTP), an artificial intelligence robot, operating a treatment planning system (TPS). Using deep reinforcement learning guided by human knowledge, we trained the VTP to autonomously adjust relevant parameters in treatment plan optimization, similar to a human planner, to generate high-quality plans for prostate cancer stereotactic body radiation therapy (SBRT). This study describes the clinical implementation and evaluation of VTP.

MATERIALS AND METHODS

We integrate VTP with Eclipse TPS using scripting Application Programming Interface. VTP observes dose-volume histograms of relevant structures, decides how to adjust dosimetric constraints, including doses, volumes, and weighting factors, and applies the adjustments to the TPS interface to launch the optimization engine. This process continues until a high-quality plan is achieved. We evaluated VTP's performance using the prostate SBRT case from the 2016 American Association of Medical Dosimetrist/Radiosurgery Society plan study with its plan scoring system, and compared to human-generated plans submitted to the challenge. Using the same scoring system, we also compared the plan quality of 36 prostate SBRT cases (20 planned with IMRT and 16 planned with VMAT) treated at our institution for both VTP and human-generated plans.

RESULTS

In the plan study case, VTP achieved a score of 142.1/150.0, ranking the third in the competition (median 134.6). For the clinical cases, VTP achieved 110.6 ± 6.5 for 20 IMRT plans and 126.2 ± 4.7 for 16 VMAT plans, similar to scores of human-generated plans with 110.4 ± 7.0 for IMRT plans and 125.4 ± 4.4 for VMAT plans. The workflow, plan quality and planning time of VTP were reviewed to be satisfactory by experienced physicists.

CONCLUSION

We successfully implemented VTP to operate a TPS for autonomous human-like treatment planning for prostate SBRT.

Online adaptive planning methods for intensity-modulated radiotherapy

Online adaptive radiation therapy aims at adapting a patient's treatment plan to their current anatomy to account for inter-fraction variations before daily treatment delivery. As this process needs to be accomplished while the patient is immobilized on the treatment couch, it requires time-efficient adaptive planning methods to generate a quality daily treatment plan rapidly. The conventional planning methods do not meet the time requirement of online adaptive radiation therapy because they often involve excessive human intervention, significantly prolonging the planning phase. This article reviews the planning strategies employed by current commercial online adaptive radiation therapy systems, research on online adaptive planning, and artificial intelligence's potential application to online adaptive planning.

Direct incorporation of patient-specific efficacy and toxicity estimates in radiation therapy plan optimization

PURPOSE

Current radiation therapy (RT) treatment planning relies mainly on pre-defined dose-based objectives and constraints to develop plans that aim to control disease while limiting damage to normal tissues during treatment. These objectives and constraints are generally population-based, in that they are developed from the aggregate response of a broad patient population to radiation. However, correlations of new biologic markers and patient-specific factors to treatment efficacy and toxicity provide the opportunity to further stratify patient populations and develop a more individualized approach to RT planning. We introduce a novel intensity-modulated radiation therapy (IMRT) optimization strategy that directly incorporates patient-specific dose response models into the planning process. In this strategy, we integrate the concept of utility-based planning where the optimization objective is to maximize the predicted value of overall treatment utility, defined by the probability of efficacy (e.g., local control) minus the weighted sum of toxicity probabilities. To demonstrate the feasibility of the approach, we apply the strategy to treatment planning for non-small cell lung cancer (NSCLC) patients.

METHODS AND MATERIALS

We developed a prioritized approach to patient-specific IMRT planning. Using a commercial treatment planning system (TPS), we calculate dose based on an influence matrix of beamlet-dose contributions to regions-of-interest. Then, outside of the TPS, we hierarchically solve two optimization problems to generate optimal beamlet weights that can then be imported back to the TPS. The first optimization problem maximizes a patient's overall plan utility subject to typical clinical dose constraints. In this process, we facilitate direct optimization of efficacy and toxicity trade-off based on individualized dose-response models. After optimal utility is determined, we solve a secondary optimization problem that minimizes a conventional dose-based objective subject to the same clinical dose constraints as the first stage but with the addition of a constraint to maintain the optimal utility from the first optimization solution. We tested this method by retrospectively generating plans for five previously treated NSCLC patients and comparing the prioritized utility plans to conventional plans optimized with only dose metric objectives. To define a plan utility function for each patient, we utilized previously published correlations of dose to local control and grade 3-5 toxicities that include patient age, stage, microRNA levels, and cytokine levels, among other clinical factors.

RESULTS

The proposed optimization approach successfully generated RT plans for five NSCLC patients that improve overall plan utility based on personalized efficacy and toxicity models while accounting for clinical dose constraints. Prioritized utility plans demonstrated the largest average improvement in local control (16.6%) when compared to plans generated with conventional planning objectives. However, for some patients, the utility-based plans resulted in similar local control estimates with decreased estimated toxicity.

CONCLUSION

The proposed optimization approach, where the maximization of a patient's RT plan utility is prioritized over the minimization of standardized dose metrics, has the potential to improve treatment outcomes by directly accounting for variability within a patient population. The implementation of the utility-based objective function offers an intuitive, humanized approach to biological optimization in which planning trade-offs are explicitly optimized.

A personalized DVH prediction model for HDR brachytherapy in cervical cancer treatment

Purpose

Although the knowledge-based dose-volume histogram (DVH) prediction has been largely researched and applied in External Beam Radiation Therapy, it is still less investigated in the domain of brachytherapy. The purpose of this study is to develop a reliable DVH prediction method for high-dose-rate brachytherapy plans.

Method

A DVH prediction workflow combining kernel density estimation (KDE), k-nearest neighbor (kNN), and principal component analysis (PCA) was proposed. PCA and kNN were first employed together to select similar patients based on principal component directions. 79 cervical cancer patients with different applicators inserted was included in this study. The KDE model was built based on the relationship between distance-to-target (DTH) and the dose in selected cases, which can be subsequently used to estimate the dose probability distribution in the validation set. Model performance of bladder and rectum was quantified by |ΔD_2cc|, |ΔD_1cc|, |ΔD_0.1cc|, |ΔD_max|, and |ΔD_mean| in the form of mean and standard deviation. The model performance between KDE only and the combination of kNN, PCA, and KDE was compared.

Result

20, 30 patients were selected for rectum and bladder based on KNN and PCA, respectively. The absolute residual between the actual plans and the predicted plans were 0.38 ± 0.29, 0.4 ± 0.32, 0.43 ± 0.36, 0.97 ± 0.66, and 0.13 ± 0.99 for |ΔD_2cc|, |ΔD_1cc|, |ΔD_0.1cc|, |ΔD_max|, and |ΔD_mean| in the bladder, respectively. For rectum, the corresponding results were 0.34 ± 0.27, 0.38 ± 0.33, 0.63 ± 0.57, 1.41 ± 0.99 and 0.23 ± 0.17, respectively. The combination of kNN, PCA, and KDE showed a significantly better prediction performance than KDE only, with an improvement of 30.3% for the bladder and 33.3% for the rectum.

Conclusion

In this study, a knowledge-based machine learning model was proposed and verified to accurately predict the DVH for new patients. This model is proved to be effective in our testing group in the workflow of HDR brachytherapy.

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.

Automatic treatment planning for cervical cancer radiation therapy using direct three-dimensional patient anatomy match

Purpose

Current knowledge‐based planning methods for radiation therapy mainly use low‐dimensional features extracted from contoured structures to identify geometrically similar patients. Here, we propose a knowledge‐based treatment planning method where the anatomical similarity is quantified by the rigid registration of the three‐dimensional (3D) planning target volume (PTV) and organs at risks (OARs) between an incoming patient and database patients.

Methods

A database that contains PTV and OARs contours from 81 cervical cancer radiation therapy patients was established. To identify the anatomically similar patients, the PTV of the new patient was registered to each PTV in the database and the Dice similarity coefficients were calculated for the PTV, rectum, and bladder between the new patient and database patients. Then the top 20 patients in the PTV match and top 3 patients in the subsequent bladder or rectum match were selected. The best dose–volume histogram parameters from the top three patients were applied as the dose constraints to the automatic plan optimization. A fast Fourier transform algorithm was developed to accelerate the 3D PTV registration process run through the database. The entire treatment planning process was automated using in‐house customized Pinnacle scripts. The automatic plans were generated for 20 patients using leave‐one‐out scheme and were evaluated against the corresponding clinical plans.

Results

The automatic plans significantly reduced rectum and bladder V50Gy by 11.79% ± 5.2% (p < 0.01) and 2.85% ± 3.16% (p < 0.01), respectively. The dose parameters achieved for the PTV and other OARs were comparable to those in the clinical plans. The entire planning process, including both dose prediction and inverse optimization, costs about 6 min.

Conclusions

The direct 3D contour match method utilizes the full spatial information of the PTV and OARs of interest and provides an intuitive measurement for patient plan anatomy similarity. The proposed automatic planning method can generate plans with better quality and higher efficiency.

A hyperparameter-tuning approach to automated inverse planning

BACKGROUND

In current practice, radiotherapy inverse planning often requires treatment planners to modify multiple parameters in the treatment planning system's objective function to produce clinically acceptable plans. Due to the manual steps in this process, plan quality can vary depending on the planning time available and the planner's skills.

PURPOSE

This study investigates the feasibility of two hyperparameter-tuning methods for automated inverse planning. Because this framework does not train a model on previously-optimized plans, it can be readily adapted to practice pattern changes, and the resulting plan quality is not limited by that of a training cohort.

METHODS

We retrospectively selected 10 patients who received lung stereotactic body radiation therapy using manually-generated clinical plans. We implemented random sampling and Bayesian optimization to automatically tune objective function parameters using linear-quadratic utility functions based on 11 clinical goals. Normalizing all plans to deliver a minimum dose of 48 Gy to 95% of the planning target volume, we compared plan quality for the automatically-generated plans to the manually-generated plans. We also investigated the impact of iteration count on the automatically-generated plans, comparing planning time and plan utility for randomized and Bayesian plans with and without stopping criteria.

RESULTS

Without stopping criteria, the median planning time was 1.9 and 2.3 hours for randomized and Bayesian plans, respectively. The organ-at-risk doses in the randomized and Bayesian plans had a median percent difference (MPD) of 48.7% and 60.4% below clinical dose limits and an MPD of 2.8% and 3.3% below clinical plan doses. With stopping criteria, the utility decreased by an MPD of 5.3% and 3.9% for randomized and Bayesian plans, but the median planning time was reduced to 0.5 and 0.7 hours, and the organ-at-risk doses still had an MPD of 42.9% and 49.7% below clinical dose limits and an MPD of 0.3% and 1.8% below clinical plan doses.

CONCLUSIONS

This study demonstrates that hyperparameter-tuning approaches to automated inverse planning can reduce the treatment planner's active planning time with plan quality that is similar to or better than manually-generated plans. This article is protected by copyright. All rights reserved.

Automatic Inverse Treatment Planning of Gamma Knife Radiosurgery via Deep Reinforcement Learning

PURPOSE

Several inverse planning algorithms have been developed for Gamma Knife (GK) radiosurgery to determine a large number of plan parameters via solving an optimization problem, which typically consists of multiple objectives. The priorities among these objectives need to be repetitively adjusted to achieve a clinically good plan for each patient. This study aimed to achieve automatic and intelligent priority-tuning, by developing a deep reinforcement learning (DRL) based method to model the tuning behaviors of human planners.

METHODS

We built a priority-tuning policy network using deep convolutional neural networks. Its input was a vector composed of multiple plan metrics that were used in our institution for GK plan evaluation. The network can determine which tuning action to take, based on the observed quality of the intermediate plan. We trained the network using an end-to-end DRL framework to approximate the optimal action-value function. A scoring function was designed to measure the plan quality to calculate the received reward of a tuning action.

RESULTS

Vestibular schwannoma was chosen as the test bed in this study. The number of training, validation and testing cases were 5, 5, and 16, respectively. For these three datasets, the average scores of the initial plans obtained with a same initial priority set were 3.63 ± 1.34, 3.83 ± 0.86 and 4.20 ± 0.78, respectively, while can be improved to 5.28 ± 0.23, 4.97 ± 0.44 and 5.22 ± 0.26 through manual priority tuning by human expert planners. Our network achieved competitive results with 5.42 ± 0.11, 5.10 ± 0. 42, 5.28 ± 0.20, respectively.

CONCLUSIONS

Our network can generate GK plans of comparable or slightly higher quality comparing with the plans generated by human planners via manual priority tuning for vestibular schwannoma cases. The network can potentially be incorporated into the clinical workflow as a planning assistance to improve GK planning efficiency and help to reduce plan quality variation caused by inter-planner variability. We also hope that our method can reduce the workload of GK planners and allow them to spend more time on more challenging cases. This article is protected by copyright. All rights reserved.

Probabilistic Pareto plan generation for semiautomated multicriteria radiation therapy treatment planning

Objective.We propose a semiautomatic pipeline for radiation therapy treatment planning, combining ideas from machine learning-automated planning and multicriteria optimization (MCO).Approach.Using knowledge extracted from historically delivered plans, prediction models for spatial dose and dose statistics are trained and furthermore systematically modified to simulate changes in tradeoff priorities, creating a set of differently biased predictions. Based on the predictions, an MCO problem is subsequently constructed using previously developed dose mimicking functions, designed in such a way that its Pareto surface spans the range of clinically acceptable yet realistically achievable plans as exactly as possible. The result is an algorithm outputting a set of Pareto optimal plans, either fluence-based or machine parameter-based, which the user can navigate between in real time to make adjustments before a final deliverable plan is created.Main results.Numerical experiments performed on a dataset of prostate cancer patients show that one may often navigate to a better plan than one produced by a single-plan-output algorithm.Significance.We demonstrate the potential of merging MCO and a data-driven workflow to automate labor-intensive parts of the treatment planning process while maintaining a certain extent of manual control for the user.

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.

Deep reinforcement learning for treatment planning in high-dose-rate cervical brachytherapy

PURPOSE

High-dose-rate (HDR) brachytherapy (BT) is an effective cancer treatment method in which the radiation source is placed within the body. Treatment planning is a critical component for a successful outcome. Almost all currently proposed treatment planning methods are built on stochastic heuristic algorithms, which limits the generation of higher quality plans. This study proposed a novel treatment planning method to adjust dwell times in a human-like fashion to improve the quality of the plan.

METHODS

We built an intelligent treatment planner network (ITPN) based on deep reinforcement learning (DRL). The network architecture of ITPN is Dueling Double-Deep Q Network. The state is the dwell time of each dwell position and the action is which dwell time to adjust and how to adjust it. A hybrid equivalent uniform dose objective function was established and assigned corresponding rewards according to its changes. Experience replay was performed with the epsilon greedy algorithm and SumTree data structure.

RESULTS

In the evaluation of ITPN using 20 patient cases, D₉₀, D₁₀₀ and V₁₀₀ showed no significant difference compared with inverse planning simulated annealing (IPSA) optimization. However, D_2cc of bladder, rectum and sigmoid, V₁₅₀ and V₂₀₀ were significant reduced, and homogeneity index and conformity index were significantly increased.

CONCLUSION

The proposed ITPN was able to generate higher quality plans based on the learned dwell time adjustment policy than IPSA. This is the first artificial intelligence system that can directly determine the dwell times of HDR BT, which demonstrated the potential feasibility of solving optimization problems via DRL.

Outcome-based multiobjective optimization of lymphoma radiation therapy plans

At its core, radiation therapy (RT) requires balancing therapeutic effects against risk of adverse events in cancer survivors. The radiation oncologist weighs numerous disease and patient-level factors when considering the expected risk-benefit ratio of combined treatment modalities. As part of this, RT plan optimization software is used to find a clinically acceptable RT plan delivering a prescribed dose to the target volume while respecting pre-defined radiation dose-volume constraints for selected organs at risk. The obvious limitation to the current approach is that it is virtually impossible to ensure the selected treatment plan could not be bettered by an alternative plan providing improved disease control and/or reduced risk of adverse events in this individual. Outcome-based optimization refers to a strategy where all planning objectives are defined by modeled estimates of a specific outcome's probability. Noting that various adverse events and disease control are generally incommensurable, leads to the concept of a Pareto-optimal plan: a plan where no single objective can be improved without degrading one or more of the remaining objectives. Further benefits of outcome-based multiobjective optimization are that quantitative estimates of risks and benefit are obtained as are the effects of choosing a different trade-off between competing objectives. Furthermore, patient-level risk factors and combined treatment modalities may be integrated directly into plan optimization. Here, we present this approach in the clinical setting of multimodality therapy for malignant lymphoma, a malignancy with marked heterogeneity in biology, target localization, and patient characteristics. We discuss future research priorities including the potential of artificial intelligence.

Machine learning for dose-volume histogram based clinical decision-making support system in radiation therapy plans for brain tumors

•

Extraction, analysis, and interpretation of historical treatment planning data is valuable but very time-consuming.

•

Proposed machine learning model classifies radiotherapy plans based on their treatment planning objectives and trade-offs.

•

Application of double nested cross-validation enabled to build a robust model that achieved 94% accuracy on a testing data.

•

Model reasoning investigated with SHAP values showed consistency with clinical observations.

Purpose

To create and investigate a novel, clinical decision-support system using machine learning (ML).

Methods and Materials

The ML model was developed based on 79 radiotherapy plans of brain tumor patients that were prescribed a total dose of 60 Gy delivered with volumetric-modulated arc therapy (VMAT). Structures considered for analysis included planning target volume (PTV), brainstem, cochleae, and optic chiasm. The model aimed to classify the target variable that included class-0 corresponding to plans for which the PTV treatment planning objective was met and class-1 that was associated with plans for which the PTV objective was not met due to the priority trade-off to meet one or more organs-at-risk constraints. Several models were evaluated using double-nested cross-validation and an area-under-the-curve (AUC) metric, with the highest performing one selected for further investigation. The model predictions were explained with Shapely additive explanation (SHAP) interaction values.

Results

The highest-performing model was Logistic Regression achieving an accuracy of 93.8 ± 4.1% and AUC of 0.98 ± 0.02 on the testing data. The SHAP analysis indicated that the ΔD_99% metric for PTV had the greatest influence on the model predictions. The least important feature was ΔD_MAX for the left and right cochleae.

Conclusions

The trained model achieved satisfactory accuracy and can be used by medical physicists in a data-driven quality assurance program as well as by radiation oncologists to support their decision-making process in terms of treatment plan approval and potential plan modifications. Model explanation analysis showed that the model relies on clinically valid logic when making predictions.

Probabilistic feature extraction, dose statistic prediction and dose mimicking for automated radiation therapy treatment planning

PURPOSE

We propose a general framework for quantifying predictive uncertainties of dose-related quantities and leveraging this information in a dose mimicking problem in the context of automated radiation therapy treatment planning.

METHODS

A three-step pipeline, comprising feature extraction, dose statistic prediction and dose mimicking, is employed. In particular, the features are produced by a convolutional variational autoencoder and used as inputs in a previously developed nonparametric Bayesian statistical method, estimating the multivariate predictive distribution of a collection of predefined dose statistics. Specially developed objective functions are then used to construct a probabilistic dose mimicking problem based on the produced distributions, creating deliverable treatment plans.

RESULTS

The numerical experiments are performed using a dataset of 94 retrospective treatment plans of prostate cancer patients. We show that the features extracted by the variational autoencoder capture geometric information of substantial relevance to the dose statistic prediction problem and are related to dose statistics in a more regularized fashion than hand-crafted features. The estimated predictive distributions are reasonable and outperforms a non-input-dependent benchmark method, and the deliverable plans produced by the probabilistic dose mimicking agree better with their clinical counterparts than for a non-probabilistic formulation.

CONCLUSIONS

We demonstrate that prediction of dose-related quantities may be extended to include uncertainty estimation and that such probabilistic information may be leveraged in a dose mimicking problem. The treatment plans produced by the proposed pipeline resemble their original counterparts well, illustrating the merits of a holistic approach to automated planning based on probabilistic modeling.

A hierarchical deep reinforcement learning framework for intelligent automatic treatment planning of prostate cancer intensity modulated radiation therapy

Purpose.We have previously proposed an intelligent automatic treatment planning (IATP) framework that builds a virtual treatment planner network (VTPN) to operate a treatment planning system (TPS) to generate high-quality radiation therapy (RT) treatment plans. While the potential of IATP in automating RT treatment planning has been demonstrated, its poor scalability caused by an almost linear growth of network size with the number of treatment planning parameters (TPPs) is a bottleneck, preventing its application in complicate, but clinically relevant treatment planning problems. The decision-making behavior of the trained network is hard to understand. Motivated by the decision-making process of a human planner, this study proposes a hierarchical IATP framework.Methods and materials.The hierarchical VTPN (HieVTPN) consists of three networks, i.e. Structure-Net, Parameter-Net, and Action-Net. When interacting with a TPS, the networks are employed in a sequential order in each step to decide the structure to adjust, the TPP to adjust for the selected structure, and the specific adjustment manner for the parameter, respectively. We developed an end-to-end hierarchical deep reinforcement learning scheme to simultaneously train the three networks. We then evaluated the effectiveness of the proposed framework in the treatment planning problems for prostate cancer intensity modulated RT (IMRT) and stereotactic body RT (SBRT). We benchmarked the performance of our approach by comparing plans made by VTPN of a parallel architecture, and the human plans submitted for competition in the 2016 American Association of Medical Dosimetrist (AAMD)/Radiosurgery Society (RSS) Plan Study. We analyzed scalability of the network size with respect to the number of TPPs. Numerical experiments were also performed to understand the rationale of the decision-making behaviors of the trained HieVTPN.Results.Both HieVTPNs for prostate IMRT and SBRT were trained successfully using 10 training patient cases and 5 validation cases. For IMRT, HieVTPN was able to generate high-quality plans for 59 testing patient cases that were not included in training process, achieving an average plan score of 8.62 (±0.83), with 9 being the maximal score. The score was comparable to that of the VTPN, 8.45 (±0.48). For SBRT planning, HieVTPN achieved an average plan score of 139.07 on five testing patient cases compared to the score of 132.21 averaged over the human plans summited for competition in AAMD/RSS plan study. Different from VTPN with network size linearly scaling with the number of TPPs, the network size of HieVTPN is almost independent of the number of TPPs. It was also observed that the decision-making behaviors of HieVTPN were understandable and generally agreed with the human experience.Conclusions.With the scalability and explainability, the hierarchical IATP framework is more favorable than the previous framework in terms of handling treatment planning problems involving a large number of TPPs.

A Web-based Dose-volume Histogram Dashboard for Library-based Individualized Dose-constraints and Clinical Plan Evaluation

Traditional methods of treatment planning and plan evaluation involve the use of generic dose-constraints. We aimed to build a web-based application to generate individualized dose-constraints and plan evaluation against a library of prior approved plan dose-volume histograms (DVH).A prototype was built for intensity modulated radiation therapy (IMRT) plans for prostate cancer. Using exported DVH files from the Varian and Accuray treatment planning systems, a library of plan DVHs was built by data extraction. Given structure volumes of a patient to be planned, a web based application was built to derive individual dose-constraints of the planning target volume (PTV) and organs-at-risk (OAR) based on achieved doses in a library of prior approved plans with similar anatomical volumes, selected using an interactive dashboard. A second web application was built to compare the achieved DVHs of the newly created plan against a library of plans of similar patients.These web application prototypes are a proof of principle that simple freely available tools can be built for library based planning and review.

A cooperative game solution approach for intensity modulated radiation therapy design: Nash Social Welfare optimization

We study the fluency map optimization problem in Intensity Modulated Radiation Therapy (IMRT) from a cooperative game theory point of view. We consider the cancerous and healthy organs in a patient's body as players of a game, where cancerous organs seek to eliminate the cancerous cells and healthy organs seek to receive no harm. The goal is to balance the trade-offs between the utility of players by forming a grand coalition between them. We do so by proposing a methodology that solves a few convex optimization problems in order to transform the fluency map optimization problem into a bargaining game. To solve the bargaining game, we employ the concept of Nash Social Welfare (NSW) optimization due to the desirable efficiency and fairness properties of its outcomes. The proposed NSW optimization is convex and can be solved by powerful commercial solvers such as CPLEX. An additional advantage of the proposed approach is that it has a new control lever for the fluency map optimization, the so-called negotiation powers, which enables practitioners to put more emphasis on an organ by changing its negotiation power. To show the efficacy of our proposed methodology, we apply it to the TG-119 case and a liver case. We compare our proposed approach with a state-of-the-art approach through creating Dose Volume Histograms.

Simultaneous needle catheter selection and dwell time optimization for preplanning of high-dose-rate brachytherapy of prostate cancer

PURPOSE

Needle catheter positions critically affect the quality of treatment plans in prostate cancer high-dose-rate (HDR) brachytherapy. The current standard needle positioning approach is based on human intuition, which cannot guarantee a high-quality plan. This study proposed a method to simultaneously select needle catheter positions and determine dwell time for preplanning of HDR brachytherapy of prostate cancer.

METHODS

We formulated the needle catheter selection problem and inverse dwell time optimization problem in a unified framework. In addition to the dose objectives of the planning target volume (PTV) and organs at risk (OARs), the objective function incorporated a group-sparsity term with a needle-specific adaptive weighting scheme to generate high-quality plans with the minimal number of needle catheters. The optimization problem was solved by a fast-iterative shrinkage-thresholding algorithm. For validation purposes, we tested the proposed algorithm on 10 patient cases previously treated at our institution and compared the resulting plans with plans generated using needle catheters selected manually.

RESULTS

Compared to the plan with manually selected needle catheters, when normalizing both plans to the same PTV coverage V _100% = 95%, the plans generated by the proposed algorithm reduced median V _125% from 65% to 64%, but increased median V _150% from 35% to 38%, and V _200% from 14% to 16%. All planning objectives were met. All clinically important dosimetric parameters of OARs were reduced. D _1cc of bladder and rectum were reduced from 8.57 Gy to 8.50 Gy and from 7.24 Gy to 6.80 Gy, respectively. D _max of urethra was reduced from 15.85 Gy to 15.77 Gy. The median number of selected needle catheters was reduced by two. The computational time for solving the proposed optimization problem was ∼90 s using MATLAB.

CONCLUSION

The proposed algorithm was able to generate plans for prostate cancer HDR brachytherapy preplanning with increased median conformity index (0.73-0.77) and slightly lower median homogeneity index (0.64-0.62) with the number of selected needles reduced by two compared to the manual needle selection approach.

Improving efficiency of training a virtual treatment planner network via knowledge-guided deep reinforcement learning for intelligent automatic treatment planning of radiotherapy

PURPOSE

We previously proposed an intelligent automatic treatment planning framework for radiotherapy, in which a virtual treatment planner network (VTPN) is built using deep reinforcement learning (DRL) to operate a treatment planning system (TPS) by adjusting treatment planning parameters in it to generate high-quality plans. We demonstrated the potential feasibility of this idea in prostate cancer intensity-modulated radiation therapy (IMRT). Despite the success, the process to train a VTPN via the standard DRL approach with an ϵ-greedy algorithm was time-consuming. The required training time was expected to grow with the complexity of the treatment planning problem, preventing the development of VTPN for more complicated but clinically relevant scenarios. In this study, we proposed a novel knowledge-guided DRL (KgDRL) approach that incorporated knowledge from human planners to guide the training process to improve the efficiency of training a VTPN.

METHOD

Using prostate cancer IMRT as a test bed, we first summarized a number of rules in the actions of adjusting treatment planning parameters of our in-house TPS. During the training process of VTPN, in addition to randomly navigating the large state-action space, as in the standard DRL approach using the ϵ-greedy algorithm, we also sampled actions defined by the rules. The priority of sampling actions from rules decreased over the training process to encourage VTPN to explore new policy on parameter adjustment that were not covered by the rules. To test this idea, we trained a VTPN using KgDRL and compared its performance with another VTPN trained using the standard DRL approach. Both networks were trained using 10 training patient cases and five additional cases for validation, while another 59 cases were employed for the evaluation purpose.

RESULTS

It was found that both VTPNs trained via KgDRL and standard DRL spontaneously learned how to operate the in-house TPS to generate high-quality plans, achieving plan quality scores of 8.82 (±0.29) and 8.43 (±0.48), respectively. Both VTPNs outperformed treatment planning purely based on the rules, which had a plan score of 7.81 (±1.59). VTPN trained with eight episodes using KgDRL was able to perform similar to that trained using DRL with 100 epochs. The training time was reduced from more than a week to ~13 hrs.

CONCLUSION

The proposed KgDRL framework was effective in accelerating the training process of a VTPN by incorporating human knowledge, which will facilitate the development of VTPN for more complicated treatment planning scenarios.

Prediction of multi-criteria optimization (MCO) parameter efficiency in volumetric modulated arc therapy (VMAT) treatment planning using machine learning (ML)

PURPOSE

To predict the impact of optimization parameter changes on dosimetric plan quality criteria in multi-criteria optimized volumetric-modulated-arc therapy (VMAT) planning prior to optimization using machine learning (ML).

METHODS

A data base comprising a total of 21,266 VMAT treatment plans for 44 cranial and 18 spinal patient geometries was generated. The underlying optimization algorithm is governed by three highly composite parameters which model a combination of important aspects of the solution. Patient geometries were parametrized via volume- and shape properties of the voxel objects and overlap-volume histograms (OVH) of the planning-target-volume (PTV) and a relevant organ-at-risk (OAR). The impact of changes in one of the three optimization parameters on the maximally achievable value range of five dosimetric properties of the resulting dose distributions was studied. To predict the extent of this impact based on patient geometry, treatment site, and current parameter settings prior to optimization, three different ML-models were trained and tested. Precision-recall curves, as well as the area-under-curve (AUC) of the resulting receiver-operator-characteristic (ROC) curves were analyzed for model assessment.

RESULTS

Successful identification of parameter regions resulting in a high variability of dosimetric plan properties depended on the choice of geometry features, the treatment indication and the plan property under investigation. AUC values between 0.82 and 0.99 could be achieved. The best average-precision (AP) values obtained from the corresponding precision/recall curves ranged from 0.71 to 0.99.

CONCLUSIONS

Machine learning models trained on a database of pre-optimized treatment plans can help finding relevant optimization parameter ranges prior to optimization.

Direct optimization of dose-volume histogram metrics in radiation therapy treatment planning

We present a method of directly optimizing on deviations in clinical goal values in radiation therapy treatment planning. Using a new mathematical framework in which metrics derived from the dose-volume histogram are regarded as functionals of an auxiliary random variable, we are able to obtain volume-at-dose and dose-at-volume as infinitely differentiable functions of the dose distribution with easily evaluable function values and gradients. Motivated by the connection to risk measures in finance, which is formalized in this framework, we also derive closed-form formulas for mean-tail-dose and demonstrate its capability of reducing extreme dose values in tail distributions. Numerical experiments performed on a prostate and a head-and-neck patient case show that the direct optimization of dose-volume histogram metrics produced marginally better results than or outperformed conventional planning objectives in terms of clinical goal fulfilment, control of low- and high-dose tails of target distributions and general plan quality defined by a pre-specified evaluation measure. The proposed framework eliminates the disconnect between optimization functions and evaluation metrics and may thus reduce the need for repetitive user interaction associated with conventional treatment planning. The method also has the potential of enhancing plan optimization in other settings such as multicriteria optimization and automated treatment planning.

Predicting patient specific Pareto fronts from patient anatomy only

PURPOSE

To demonstrate the feasibility of predicting the patient-specific treatment planning Pareto front (PF) for prostate cancer patients based only on delineations of PTV, rectum and body.

MATERIAL/METHODS

Our methodology consists of four steps. First, using Erasmus-iCycle, the Pareto fronts of 112 prostate cancer patients were constructed by generating per patient 42 Pareto optimal treatment plans with different priorities. Dose parameters associated to homogeneity, conformity and dose to rectum were extracted. Second, a 3D convex function representing the PF spanned by the 42 plans was fitted for each patient using three patient-specific parameters. Third, ten features were extracted from the, aforementioned, structures to train a linear-regressor prediction algorithm to predict these three patient-specific parameters. Fourth, the quality of the predictions was assessed by calculating the average and maximum distances of the predicted PF to the 42 plans for patients in the validation cohort.

RESULTS

The prediction model was able to predict the clinically relevant PF within 2 Gy for 90% of the patients with a median average distance of 0.6 Gy.

CONCLUSIONS

We demonstrate the feasibility of fast, accurate predictions of the patient-specific PF for prostate cancer patients based only on delineations of PTV, rectum and body.

Artificial intelligence in radiation oncology

Artificial intelligence (AI) has the potential to fundamentally alter the way medicine is practised. AI platforms excel in recognizing complex patterns in medical data and provide a quantitative, rather than purely qualitative, assessment of clinical conditions. Accordingly, AI could have particularly transformative applications in radiation oncology given the multifaceted and highly technical nature of this field of medicine with a heavy reliance on digital data processing and computer software. Indeed, AI has the potential to improve the accuracy, precision, efficiency and overall quality of radiation therapy for patients with cancer. In this Perspective, we first provide a general description of AI methods, followed by a high-level overview of the radiation therapy workflow with discussion of the implications that AI is likely to have on each step of this process. Finally, we describe the challenges associated with the clinical development and implementation of AI platforms in radiation oncology and provide our perspective on how these platforms might change the roles of radiotherapy medical professionals.

Dr. Answer AI for prostate cancer: Clinical outcome prediction model and service

Objectives

The importance of clinical outcome prediction models using artificial intelligence (AI) is being emphasized owing to the increasing necessity of developing a clinical decision support system (CDSS) employing AI. Therefore, in this study, we proposed a “Dr. Answer” AI software based on the clinical outcome prediction model for prostate cancer treated with radical prostatectomy.

Methods

The Dr. Answer AI was developed based on a clinical outcome prediction model, with a user-friendly interface. We used 7,128 clinical data of prostate cancer treated with radical prostatectomy from three hospitals. An outcome prediction model was developed to calculate the probability of occurrence of 1) tumor, node, and metastasis (TNM) staging, 2) extracapsular extension, 3) seminal vesicle invasion, and 4) lymph node metastasis. Random forest and k-nearest neighbors algorithms were used, and the proposed system was compared with previous algorithms.

Results

Random forest exhibited good performance for TNM staging (recall value: 76.98%), while k-nearest neighbors exhibited good performance for extracapsular extension, seminal vesicle invasion, and lymph node metastasis (80.24%, 98.67%, and 95.45%, respectively). The Dr. Answer AI software consisted of three primary service structures: 1) patient information, 2) clinical outcome prediction, and outcomes according to the National Comprehensive Cancer Network guideline.

Conclusion

The proposed clinical outcome prediction model could function as an effective CDSS, supporting the decisions of the physicians, while enabling the patients to understand their treatment outcomes. The Dr. Answer AI software for prostate cancer helps the doctors to explain the treatment outcomes to the patients, allowing the patients to be more confident about their treatment plans.

Operating a treatment planning system using a deep-reinforcement learning-based virtual treatment planner for prostate cancer intensity-modulated radiation therapy treatment planning

PURPOSE

In the treatment planning process of intensity-modulated radiation therapy (IMRT), a human planner operates the treatment planning system (TPS) to adjust treatment planning parameters, for example, dose volume histogram (DVH) constraints' locations and weights, to achieve a satisfactory plan for each patient. This process is usually time-consuming, and the plan quality depends on planer's experience and available planning time. In this study, we proposed to model the behaviors of human planners in treatment planning by a deep reinforcement learning (DRL)-based virtual treatment planner network (VTPN), such that it can operate the TPS in a human-like manner for treatment planning.

METHODS AND MATERIALS

Using prostate cancer IMRT as an example, we established the VTPN using a deep neural network developed. We considered an in-house optimization engine with a weighted quadratic objective function. Virtual treatment planner network was designed to observe an intermediate plan DVHs and decide the action to improve the plan by changing weights and threshold dose in the objective function. We trained the VTPN in an end-to-end DRL process in 10 patient cases. A plan score was used to measure plan quality. We demonstrated the feasibility and effectiveness of the trained VTPN in another 64 patient cases.

RESULTS

Virtual treatment planner network was trained to spontaneously learn how to adjust treatment planning parameters to generate high-quality treatment plans. In the 64 testing cases, with initialized parameters, quality score was 4.97 (±2.02), with 9.0 being the highest possible score. Using VTPN to perform treatment planning improved quality score to 8.44 (±0.48).

CONCLUSIONS

To our knowledge, this was the first time that intelligent treatment planning behaviors of human planner in external beam IMRT are autonomously encoded in an artificial intelligence system. The trained VTPN is capable of behaving in a human-like way to produce high-quality plans.

Artificial Intelligence in Radiotherapy Treatment Planning: Present and Future

Treatment planning is an essential step of the radiotherapy workflow. It has become more sophisticated over the past couple of decades with the help of computer science, enabling planners to design highly complex radiotherapy plans to minimize the normal tissue damage while persevering sufficient tumor control. As a result, treatment planning has become more labor intensive, requiring hours or even days of planner effort to optimize an individual patient case in a trial-and-error fashion. More recently, artificial intelligence has been utilized to automate and improve various aspects of medical science. For radiotherapy treatment planning, many algorithms have been developed to better support planners. These algorithms focus on automating the planning process and/or optimizing dosimetric trade-offs, and they have already made great impact on improving treatment planning efficiency and plan quality consistency. In this review, the smart planning tools in current clinical use are summarized in 3 main categories: automated rule implementation and reasoning, modeling of prior knowledge in clinical practice, and multicriteria optimization. Novel artificial intelligence-based treatment planning applications, such as deep learning-based algorithms and emerging research directions, are also reviewed. Finally, the challenges of artificial intelligence-based treatment planning are discussed for future works.

Intelligent inverse treatment planning via deep reinforcement learning, a proof-of-principle study in high dose-rate brachytherapy for cervical cancer

Inverse treatment planning in radiation therapy is formulated as solving optimization problems. The objective function and constraints consist of multiple terms designed for different clinical and practical considerations. Weighting factors of these terms are needed to define the optimization problem. While a treatment planning optimization engine can solve the optimization problem with given weights, adjusting the weights to yield a high-quality plan is typically performed by a human planner. Yet the weight-tuning task is labor intensive, time consuming, and it critically affects the final plan quality. An automatic weight-tuning approach is strongly desired. The procedure of weight adjustment to improve the plan quality is essentially a decision-making problem. Motivated by the tremendous success in deep learning for decision making with human-level intelligence, we propose a novel framework to adjust the weights in a human-like manner. This study used inverse treatment planning in high-dose-rate brachytherapy (HDRBT) for cervical cancer as an example. We developed a weight-tuning policy network (WTPN) that observes dose volume histograms of a plan and outputs an action to adjust organ weighting factors, similar to the behaviors of a human planner. We trained the WTPN via end-to-end deep reinforcement learning. Experience replay was performed with the epsilon greedy algorithm. After training was completed, we applied the trained WTPN to guide treatment planning of five testing patient cases. It was found that the trained WTPN successfully learnt the treatment planning goals and was able to guide the weight tuning process. On average, the quality score of plans generated under the WTPN's guidance was improved by ~8.5% compared to the initial plan with arbitrarily set weights, and by 10.7% compared to the plans generated by human planners. To our knowledge, this was the first time that a tool was developed to adjust organ weights for the treatment planning optimization problem in a human-like fashion based on intelligence learnt from a training process, which was different from existing strategies based on pre-defined rules. The study demonstrated potential feasibility to develop intelligent treatment planning approaches via deep reinforcement learning.

Knowledge-based planning for intensity-modulated radiation therapy: A review of data-driven approaches

Purpose

Intensity‐Modulated Radiation Therapy (IMRT), including its variations (including IMRT, Volumetric Arc Therapy (VMAT), and Tomotherapy), is a widely used and critically important technology for cancer treatment. It is a knowledge‐intensive technology due not only to its own technical complexity, but also to the inherently conflicting nature of maximizing tumor control while minimizing normal organ damage. As IMRT experience and especially the carefully designed clinical plan data are accumulated during the past two decades, a new set of methods commonly termed knowledge‐based planning (KBP) have been developed that aim to improve the quality and efficiency of IMRT planning by learning from the database of past clinical plans. Some of this development has led to commercial products recently that allowed the investigation of KBP in numerous clinical applications. In this literature review, we will attempt to present a summary of published methods of knowledge‐based approaches in IMRT and recent clinical validation results.

Methods

In March 2018, a literature search was conducted in the NIH Medline database using the PubMed interface to identify publications that describe methods and validations related to KBP in IMRT including variations such as VMAT and Tomotherapy. The search criteria were designed to have a broad scope to capture relevant results with high sensitivity. The authors filtered down the search results according to a predefined selection criteria by reviewing the titles and abstracts first and then by reviewing the full text. A few papers were added to the list based on the references of the reviewed papers. The final set of papers was reviewed and summarized here.

Results

The initial search yielded a total of 740 articles. A careful review of the titles, abstracts, and eventually the full text and then adding relevant articles from reviewing the references resulted in a final list of 73 articles published between 2011 and early 2018. These articles described methods for developing knowledge models for predicting such parameters as dosimetric and dose‐volume points, voxel‐level doses, and objective function weights that improve or automate IMRT planning for various cancer sites, addressing different clinical and quality assurance needs, and using a variety of machine learning approaches. A number of articles reported carefully designed clinical studies that assessed the performance of KBP models in realistic clinical applications. Overwhelming majority of the studies demonstrated the benefits of KBP in achieving comparable and often improved quality of IMRT planning while reducing planning time and plan quality variation.

Conclusions

The number of KBP‐related studies has been steadily increasing since 2011 indicating a growing interest in applying this approach to clinical applications. Validation studies have generally shown KBP to produce plans with quality comparable to expert planners while reducing the time and efforts to generate plans. However, current studies are mostly retrospective and leverage relatively small datasets. Larger datasets collected through multi‐institutional collaboration will enable the development of more advanced models to further improve the performance of KBP in complex clinical cases. Prospective studies will be an important next step toward widespread adoption of this exciting technology.

Advanced Treatment Planning

Treatment planning for protons and heavier ions is adapting technologies originally developed for photon dose optimization, but also has to meet its particular challenges. Since the quality of the applied dose is more sensitive to geometric uncertainties, treatment plan robust optimization has a much more prominent role in particle therapy. This has led to specific planning tools, approaches, and research into new formulations of the robust optimization problems. Tools for solution space navigation and automatic planning are also being adapted to particle therapy. These challenges become even greater when detailed models of relative biological effectiveness (RBE) are included into dose optimization, as is required for heavier ions.

A small number of objective function weight vectors is sufficient for automated treatment planning in prostate cancer

Current practice for treatment planning optimization can be both inefficient and time consuming. In this paper, we propose an automated planning methodology that aims to combine both explorative and prescriptive approaches for improving the efficiency and the quality of the treatment planning process. Given a treatment plan, our explorative approach explores trade-offs between different objectives and finds an acceptable region for objective function weights via inverse optimization. Intuitively, the shape and size of these regions describe how 'sensitive' a patient is to perturbations in objective function weights. We then develop an integer programming-based prescriptive approach that exploits the information encoded by these regions to find a set of five representative objective function weight vectors such that for each patient there exists at least one representative weight vector that can produce a high quality treatment plan. Using 315 patients from Princess Margaret Cancer Centre, we show that the produced treatment plans are comparable and, for [Formula: see text] of cases, improve upon the inversely optimized plans that are generated from the historical clinical treatment plans.

Automation in intensity modulated radiotherapy treatment planning-a review of recent innovations

Radiotherapy treatment planning of complex radiotherapy techniques, such as intensity modulated radiotherapy and volumetric modulated arc therapy, is a resource-intensive process requiring a high level of treatment planner intervention to ensure high plan quality. This can lead to variability in the quality of treatment plans and the efficiency in which plans are produced, depending on the skills and experience of the operator and available planning time. Within the last few years, there has been significant progress in the research and development of intensity modulated radiotherapy treatment planning approaches with automation support, with most commercial manufacturers now offering some form of solution. There is a rapidly growing number of research articles published in the scientific literature on the topic. This paper critically reviews the body of publications up to April 2018. The review describes the different types of automation algorithms, including the advantages and current limitations. Also included is a discussion on the potential issues with routine clinical implementation of such software, and highlights areas for future research.

A deep learning-based prediction model for gamma evaluation in patient-specific quality assurance

PURPOSE

Patient-specific quality assurance (QA) measurement is conducted to confirm the accuracy of dose delivery. However, measurement is time-consuming and places a heavy workload on the medical physicists and radiological technologists. In this study, we proposed a prediction model for gamma evaluation, based on deep learning. We applied the model to a QA measurement dataset of prostate cancer cases to evaluate its practicality.

METHODS

Sixty pretreatment verification plans from prostate cancer patients treated using intensity modulated radiation therapy were collected. Fifteen-layer convolutional neural networks (CNN) were developed to learn the sagittal planar dose distributions from a RT-3000 QA phantom (R-TECH.INC., Tokyo, Japan). The percentage gamma passing rate (GPR) was measured using GAFCHROMIC EBT3 film (Ashland Specialty Ingredients, Covington, USA). The input training data also included the volume of the PTV (planning target volume), rectum, and overlapping region, measured in cm3 , and the monitor unit values for each field. The network produced predicted GPR values at four criteria: 2%(global)/2 mm, 3%(global)/2 mm, 2%(global)/3 mm, and 3%(global)/3 mm. Adam, an algorithm for first-order gradient-based optimization of stochastic objective functions, was used for learning and for optimizing the CNN-based model. Fivefold cross-validation was applied to validate the performance of the proposed method. Forty cases were used for training and validation set in fivefold cross-validation, and the remaining 20 cases were used for the test set. The predicted and measured GPR values were compared.

RESULTS

A linear relationship was found between the measured and predicted values, for each of the four criteria. Spearman rank correlation coefficients in validation set between measured and predicted GPR values at four criteria were 0.73 at 2%/2 mm, 0.72 at 3%/2 mm, 0.74 at 2%/3 mm, and 0.65 at 3%/3 mm, respectively (P < 0.01). The Spearman rank correlation coefficients in the test set were 0.62 (P < 0.01) at 2%/2 mm, 0.56 (P < 0.01) at 3%/2 mm, 0.51 (P = 0.02) at 2%/3 mm, and 0.32 (P = 0.16) at 3%/3 mm. These results demonstrated a strong or moderate correlation between the predicted and measured values.

CONCLUSIONS

We developed a CNN-based prediction model for patient-specific QA of dose distribution in prostate treatment. Our results suggest that deep learning may provide a useful prediction model for gamma evaluation of patient-specific QA in prostate treatment planning.

Sample size requirements for knowledge-based treatment planning

PURPOSE

To determine how training set size affects the accuracy of knowledge-based treatment planning (KBP) models.

METHODS

The authors selected four models from three classes of KBP approaches, corresponding to three distinct quantities that KBP models may predict: dose-volume histogram (DVH) points, DVH curves, and objective function weights. DVH point prediction is done using the best plan from a database of similar clinical plans; DVH curve prediction employs principal component analysis and multiple linear regression; and objective function weights uses either logistic regression or K-nearest neighbors. The authors trained each KBP model using training sets of sizes n = 10, 20, 30, 50, 75, 100, 150, and 200. The authors set aside 100 randomly selected patients from their cohort of 315 prostate cancer patients from Princess Margaret Cancer Center to serve as a validation set for all experiments. For each value of n, the authors randomly selected 100 different training sets with replacement from the remaining 215 patients. Each of the 100 training sets was used to train a model for each value of n and for each KBT approach. To evaluate the models, the authors predicted the KBP endpoints for each of the 100 patients in the validation set. To estimate the minimum required sample size, the authors used statistical testing to determine if the median error for each sample size from 10 to 150 is equal to the median error for the maximum sample size of 200.

RESULTS

The minimum required sample size was different for each model. The DVH point prediction method predicts two dose metrics for the bladder and two for the rectum. The authors found that more than 200 samples were required to achieve consistent model predictions for all four metrics. For DVH curve prediction, the authors found that at least 75 samples were needed to accurately predict the bladder DVH, while only 20 samples were needed to predict the rectum DVH. Finally, for objective function weight prediction, at least 10 samples were needed to train the logistic regression model, while at least 150 samples were required to train the K-nearest neighbor methodology.

CONCLUSIONS

In conclusion, the minimum required sample size needed to accurately train KBP models for prostate cancer depends on the specific model and endpoint to be predicted. The authors' results may provide a lower bound for more complicated tumor sites.