Effect of total body irradiation lung block parameters on lung doses using three-dimensional dosimetry.

PURPOSE: Total body irradiation (TBI) is an integral part of stem cell transplant. However, patients are at risk of treatment-related toxicities, including radiation pneumonitis. While lung dose is one of the most crucial aspects of TBI dosimetry, currently available data are based on point doses. As volumetric dose distribution could be substantially altered by lung block parameters, we used 3D dosimetry in our treatment planning system to estimate volumetric lung dose and measure the impact of various lung block designs. MATERIALS AND METHODS: We commissioned a TBI beam model in RayStation that matches the measured tissue-phantom ratio under our clinical TBI setup. Cerrobend blocks were automatically generated in RayStation on thoracic Computed Tomography (CT) scans from three anonymized patients using the lung, clavicle, spine, and diaphragmatic contours. The margin for block edge was varied to 0, 1, or 2 cm from the superior, lateral, and inferior thoracic borders, with a uniform margin 2.5 cm lateral to the vertebral bodies. The lung dose was calculated and compared with a prescription dose of 1200 cGy in six fractions (three with blocks and three without). RESULT: The point dose at midplane under the block and the average lung dose are at the range of 73%-76% and 80%-88% of prescription dose respectively regardless of the block margins. In contrast, the percent lung volume receiving 10 Gy increased by nearly two-fold, from 31% to 60% over the margins from 0 to 2 cm. CONCLUSIONS: The TPS-derived 3D lung dose is substantially different from the nominal dose assumed with HVL lung blocks. Point doses under the block are insufficient to accurately gauge the relationship between dose and pneumonitis, and TBI dosimetry could be highly variable between patients and institutions as more descriptive parameters are not included in protocols. Much progress remains to be made to optimize and standardize technical aspects of TBI, and better dosimetry could provide more precise dosimetric predictors for pneumonitis risk.

Validation of the Pinnacle³ photon convolution-superposition algorithm applied to fast neutron beams.

We evaluate a photon convolution-superposition algorithm used to model a fast neutron therapy beam in a commercial treatment planning system (TPS). The neutron beam modeled was the Clinical Neutron Therapy System (CNTS) fast neutron beam produced by 50 MeV protons on a Be target at our facility, and we implemented the Pinnacle3 dose calculation model for computing neutron doses. Measured neutron data were acquired by an IC30 ion chamber flowing 5 cc/min of tissue equivalent gas. Output factors and profile scans for open and wedged fields were measured according to the Pinnacle physics reference guide recommendations for photon beams in a Wellhofer water tank scanning system. Following the construction of a neutron beam model, computed doses were then generated using 100 monitor units (MUs) beams incident on a water-equivalent phantom for open and wedged square fields, as well as multileaf collimator (MLC)-shaped irregular fields. We compared Pinnacle dose profiles, central axis doses, and off-axis doses (in irregular fields) with 1) doses computed using the Prism treatment planning system, and 2) doses measured in a water phantom and having matching geometry to the computation setup. We found that the Pinnacle photon model may be used to model most of the important dosimetric features of the CNTS fast neutron beam. Pinnacle-calculated dose points among open and wedged square fields exhibit dose differences within 3.9 cGy of both Prism and measured doses along the central axis, and within 5 cGy difference of measurement in the penumbra region. Pinnacle dose point calculations using irregular treatment type fields showed a dose difference up to 9 cGy from measured dose points, although most points of comparison were below 5 cGy. Comparisons of dose points that were chosen from cases planned in both Pinnacle and Prism show an average dose difference less than 0.6%, except in certain fields which incorporate both wedges and heavy blocking of the central axis. All clinical cases planned in both Prism and Pinnacle were found to be comparable in terms of dose-volume histograms and spatial dose distribution following review by the treating clinicians. Variations were considered minor and within clinically acceptable limits by the treating clinicians. The Pinnacle TPS has sufficient computational modeling ability to adequately produce a viable neutron model for clinical use in treatment planning.

When is better best? A multiobjective perspective.

PURPOSE: To identify the most informative methods for reporting results of treatment planning comparisons. METHODS: Seven articles from the past year of International Journal of Radiation Oncology Biology Physics reported on comparisons of treatment plans for IMRT and IMAT. The articles were reviewed to identify methods of comparisons. Decision theoretical concepts were used to evaluate the study methods and highlight those that provide the most information. RESULTS: None of the studies examined the correlation between objectives. Statistical comparisons provided some information but not enough to provide support for a robust decision analysis. CONCLUSIONS: The increased use of treatment planning studies to evaluate different methods in radiation therapy requires improved standards for designing the studies and reporting the results.

Investigation of effective decision criteria for multiobjective optimization in IMRT.

PURPOSE: To investigate how using different sets of decision criteria impacts the quality of intensity modulated radiation therapy (IMRT) plans obtained by multiobjective optimization. METHODS: A multiobjective optimization evolutionary algorithm (MOEA) was used to produce sets of IMRT plans. The MOEA consisted of two interacting algorithms: (i) a deterministic inverse planning optimization of beamlet intensities that minimizes a weighted sum of quadratic penalty objectives to generate IMRT plans and (ii) an evolutionary algorithm that selects the superior IMRT plans using decision criteria and uses those plans to determine the new weights and penalty objectives of each new plan. Plans resulting from the deterministic algorithm were evaluated by the evolutionary algorithm using a set of decision criteria for both targets and organs at risk (OARs). Decision criteria used included variation in the target dose distribution, mean dose, maximum dose, generalized equivalent uniform dose (gEUD), an equivalent uniform dose (EUD(alpha,beta) formula derived from the linear-quadratic survival model, and points on dose volume histograms (DVHs). In order to quantatively compare results from trials using different decision criteria, a neutral set of comparison metrics was used. For each set of decision criteria investigated, IMRT plans were calculated for four different cases: two simple prostate cases, one complex prostate Case, and one complex head and neck Case. RESULTS: When smaller numbers of decision criteria, more descriptive decision criteria, or less anti-correlated decision criteria were used to characterize plan quality during multiobjective optimization, dose to OARs and target dose variation were reduced in the final population of plans. Mean OAR dose and gEUD (a = 4) decision criteria were comparable. Using maximum dose decision criteria for OARs near targets resulted in inferior populations that focused solely on low target variance at the expense of high OAR dose. Target dose range, (D(max) - D(min)), decision criteria were found to be most effective for keeping targets uniform. Using target gEUD decision criteria resulted in much lower OAR doses but much higher target dose variation. EUD(alpha,beta) based decision criteria focused on a region of plan space that was a compromise between target and OAR objectives. None of these target decision criteria dominated plans using other criteria, but only focused on approaching a different area of the Pareto front. CONCLUSIONS: The choice of decision criteria implemented in the MOEA had a significant impact on the region explored and the rate of convergence toward the Pareto front. When more decision criteria, anticorrelated decision criteria, or decision criteria with insufficient information were implemented, inferior populations are resulted. When more informative decision criteria were used, such as gEUD, EUD(alpha,beta), target dose range, and mean dose, MOEA optimizations focused on approaching different regions of the Pareto front, but did not dominate each other. Using simple OAR decision criteria and target EUD(alpha,beta) decision criteria demonstrated the potential to generate IMRT plans that significantly reduce dose to OARs while achieving the same or better tumor control when clinical requirements on target dose variance can be met or relaxed.

A hierarchical evolutionary algorithm for multiobjective optimization in IMRT.

PURPOSE: The current inverse planning methods for intensity modulated radiation therapy (IMRT) are limited because they are not designed to explore the trade-offs between the competing objectives of tumor and normal tissues. The goal was to develop an efficient multiobjective optimization algorithm that was flexible enough to handle any form of objective function and that resulted in a set of Pareto optimal plans. METHODS: A hierarchical evolutionary multiobjective algorithm designed to quickly generate a small diverse Pareto optimal set of IMRT plans that meet all clinical constraints and reflect the optimal trade-offs in any radiation therapy plan was developed. The top level of the hierarchical algorithm is a multiobjective evolutionary algorithm (MOEA). The genes of the individuals generated in the MOEA are the parameters that define the penalty function minimized during an accelerated deterministic IMRT optimization that represents the bottom level of the hierarchy. The MOEA incorporates clinical criteria to restrict the search space through protocol objectives and then uses Pareto optimality among the fitness objectives to select individuals. The population size is not fixed, but a specialized niche effect, domination advantage, is used to control the population and plan diversity. The number of fitness objectives is kept to a minimum for greater selective pressure, but the number of genes is expanded for flexibility that allows a better approximation of the Pareto front. RESULTS: The MOEA improvements were evaluated for two example prostate cases with one target and two organs at risk (OARs). The population of plans generated by the modified MOEA was closer to the Pareto front than populations of plans generated using a standard genetic algorithm package. Statistical significance of the method was established by compiling the results of 25 multiobjective optimizations using each method. From these sets of 12-15 plans, any random plan selected from a MOEA population had a 11.3% +/- 0.7% chance of dominating any random plan selected by a standard genetic package with 0.04% +/- 0.02% chance of domination in reverse. By implementing domination advantage and protocol objectives, small and diverse populations of clinically acceptable plans that approximated the Pareto front could be generated in a fraction of 1 h. Acceleration techniques implemented on both levels of the hierarchical algorithm resulted in short, practical runtimes for multiobjective optimizations. CONCLUSIONS: The MOEA produces a diverse Pareto optimal set of plans that meet all dosimetric protocol criteria in a feasible amount of time. The final goal is to improve practical aspects of the algorithm and integrate it with a decision analysis tool or human interface for selection of the IMRT plan with the best possible balance of successful treatment of the target with low OAR dose and low risk of complication for any specific patient situation.

Radiation Therapy Quality Assurance Tasks and Tools: The Many Roles of Machine Learning.

The recent explosion in machine learning efforts in the quality assurance (QA) space has produced a variety of proofs-of-concept many with promising results. Expected outcomes of model implementation include improvements in planning time, plan quality, advanced dosimetric QA, predictive machine maintenance, increased safety checks, and developments key for new QA paradigms driven by adaptive planning. In this article, we outline several areas of research and discuss some of the unique challenges each area presents.

Ontologies in radiation oncology.

Ontologies are a formal, computer-compatible method for representing scientific knowledge about a given domain. They provide a standardized vocabulary, taxonomy and set of relations between concepts. When formatted in a standard way, they can be read and reasoned upon by computers as well as by humans. At the 2019 International Conference on the Use of Computers in Radiation Therapy, there was a session devoted to ontologies in radiation therapy. This paper is a compilation of the material presented, and is meant as an introduction to the subject. This is done by means of a didactic introduction to the topic followed by a series of applications in radiation therapy. The goal of this article is to provide the medical physicist and related professionals with sufficient background that they can understand their construction as well as their practical uses.

Characterization of a Bayesian network-based radiotherapy plan verification model.

PURPOSE: The current process for radiotherapy treatment plan quality assurance relies on human inspection of treatment plans, which is time-consuming, error prone and oft reliant on inconsistently applied professional judgments. A previous proof-of-principle paper describes the use of a Bayesian network (BN) to aid in this process. This work studied how such a BN could be expanded and trained to better represent clinical practice. METHODS: We obtained 51 540 unique radiotherapy cases including diagnostic, prescription, plan/beam, and therapy setup factors from a de-identified Elekta oncology information system from the years 2010-2017 from a single institution. Using a knowledge base derived from clinical experience, factors were coordinated into a 29-node, 40-edge BN representing dependencies among the variables. Conditional probabilities were machine learned using expectation maximization module using all data except a subset of 500 patient cases withheld for testing. Different classes of errors that were obtained from incident learning systems were introduced to the testing set of cases which were withheld from the dataset used for building the BN. Different sizes of datasets were used to train the network. In addition, the BN was trained using data from different length epochs as well as different eras. Its performance under these different conditions was evaluated by means of Areas Under the receiver operating characteristic Curve (AUC). RESULTS: Our performance analysis found AUCs of 0.82, 0.85, 0.89, and 0.88 in networks trained with 2-yr, 3-yr 4-yr and 5-yr windows. With a 4-yr sliding window, we found AUC reduction of 3% per year when moving the window back in time in 1-yr steps. Compared to the 4-yr window moved back by 4 yrs (2010-2013 vs 2014-2017), the largest component of overall reduction in AUC over time was from the loss of detection performance in plan/beam error types. CONCLUSIONS: The expanded BN method demonstrates the ability to detect classes of errors commonly encountered in radiotherapy planning. The results suggest that a 4-yr training dataset optimizes the performance of the network in this institutional dataset, and that yearly updates are sufficient to capture the evolution of clinical practice and maintain fidelity.

Applying Probabilistic Decision Models to Clinical Trial Design.

Clinical trial design most often focuses on a single or several related outcomes with corresponding calculations of statistical power. We consider a clinical trial to be a decision problem, often with competing outcomes. Using a current controversy in the treatment of HPV-positive head and neck cancer, we apply several different probabilistic methods to help define the range of outcomes given different possible trial designs. Our model incorporates the uncertainties in the disease process and treatment response and the inhomogeneities in the patient population. Instead of expected utility, we have used a Markov model to calculate quality adjusted life expectancy as a maximization objective. Monte Carlo simulations over realistic ranges of parameters are used to explore different trial scenarios given the possible ranges of parameters. This modeling approach can be used to better inform the initial trial design so that it will more likely achieve clinical relevance.

Decision analytic modeling for the economic analysis of proton radiotherapy for non-small cell lung cancer.

BACKGROUND: Although proton radiation treatments are more costly than photon/X-ray therapy, they may lower overall treatment costs through reducing rates of severe toxicities and the costly management of those toxicities. To study this issue, we created a decision-model comparing proton vs. X-ray radiotherapy for locally advanced non-small cell lung cancer patients. METHODS: An influence diagram was created to model for radiation delivery, associated 6-month pneumonitis/esophagitis rates, and overall costs (radiation plus toxicity costs). Pneumonitis (age, chemo type, V20, MLD) and esophagitis (V60) predictors were modeled to impact toxicity rates. We performed toxicity-adjusted, rate-adjusted, risk group-adjusted, and radiosensitivity analyses. RESULTS: Upfront proton treatment costs exceeded that of photons [$16,730.37 (3DCRT), $23,893.83 (IMRT), $41,061.80 (protons)]. Based upon expected population pneumonitis and esophagitis rates for each modality, protons would be expected to recover $1,065.62 and $1,139.63 of the cost difference compared to 3DCRT or IMRT. For patients treated with IMRT experiencing grade 4 pneumonitis or grade 4 esophagitis, costs exceeded patients treated with protons without this toxicity. 3DCRT patients with grade 4 esophagitis had higher costs than proton patients without this toxicity. For the risk group analysis, high risk patients (age >65, carboplatin/paclitaxel) benefited more from proton therapy. A biomarker may allow patient selection for proton therapy, although the AUC alone is not sufficient to determine if the biomarker is clinically useful. CONCLUSIONS: The comparison between proton and photon/X-ray radiation therapy for NSCLC needs to consider both the up-front cost of treatment and the possible long term cost of complications. In our analysis, current costs favor X-ray therapy. However, relatively small reductions in the cost of proton therapy may result in a shift to the preference for proton therapy.

Developing Bayesian networks from a dependency-layered ontology: A proof-of-concept in radiation oncology.

PURPOSE: Bayesian networks (BNs) are graphical representations of probabilistic knowledge that offer normative reasoning under uncertainty and are well suited for use in medical domains. Traditional knowledge-based network development of BN topology requires that modeling experts establish relevant dependency links between domain concepts by searching and translating published literature, querying domain experts, or applying machine learning algorithms on data. For initial development these methods are time-intensive and this cost hinders the growth of BN applications in medical decision making. Further, this approach fails to utilize knowledge representation in medical fields to automate network development. Our research alleviates the challenges surrounding BN modeling in radiation oncology by leveraging an ontology based hub and spoke system for BN construction. METHODS: We implement a hub and spoke system by developing (a) an ontology of knowledge in radiation oncology (the hub) which includes dependency semantics similar to BN relations and (b) a software tool that operates on ontological semantics using deductive reasoning to create BN topologies (the spokes). We demonstrate that network topologies built using the software are terminologically consistent and form networks that are topologically compatible with existing ones. We do this first by merging two different BN models for prostate cancer radiotherapy prediction which contain domain cross terms. We then use the logic to perform discovery of new causal chains between radiation oncology concepts. RESULTS: From the radiation oncology (RO) ontology we successfully reconstructed a previously published prostate cancer radiotherapy Bayes net using up-to-date domain knowledge. Merging this model with another similar prostate cancer model in the RO domain produced a larger, highly interconnected model representing the expanded scope of knowledge available regarding prostate cancer therapy parameters, complications, and outcomes. The causal discovery resulted in an automatically-built causal network model of all ontologized radiotherapy concepts between a 'Mucositis' complication and anatomic tumor location. CONCLUSIONS: The proposed model building approach lowers barriers to developing probabilistic models relevant to real-world clinical decision making, and offers a solution to the consistency and compatibility problems. Further, the knowledge representation in this work demonstrates potential for broader radiation oncology applications outside of Bayes nets.

Clinical Positioning Accuracy for Multisession Stereotactic Radiotherapy With the Gamma Knife Perfexion.

Multisession stereotactic radiation therapy is increasingly being seen as a preferred option for intracranial diseases in close proximity to critical structures and for larger target volumes. The objective of this study is to investigate the reproducibility of the Extend system from Elekta. A retrospective review was conducted for all patients treated with multisession Gamma Knife between July 2010 and June 2015, including both malignant and benign lesions. Eighty-four patients were treated in this 5-year span. The average residual daily setup uncertainty was 0.48 (0.19) mm. We compare measurements of setup uncertainty from the Extend system to measurements performed with a linac-based approach previously used in our center. The Extend system has significantly reduced setup uncertainty for fractionated intracranial treatments at our institution. Positive results were observed in a small population of edentulous patients. The Extend system compares favorably with other approaches to delivering intracranial stereotactic radiotherapy and is a robust, simple-to-use, and precise method for treating multisession intracranial lesions.

A feasibility study of dynamic adaptive radiotherapy for nonsmall cell lung cancer.

PURPOSE: The final state of the tumor at the end of a radiotherapy course is dependent on the doses given in each fraction during the treatment course. This study investigates the feasibility of using dynamic adaptive radiotherapy (DART) in treating lung cancers assuming CBCT is available to observe midtreatment tumor states. DART adapts treatment plans using a dynamic programming technique to consider the expected changes of the tumor in the optimization process. METHODS: DART is constructed using a stochastic control formalism framework. It minimizes the total expected number of tumor cells at the end of a treatment course, which is equivalent to maximizing tumor control probability, subject to the uncertainty inherent in the tumor response. This formulation allows for nonstationary dose distributions as well as nonstationary fractional doses as needed to achieve a series of optimal plans that are conformal to the tumor over time, i.e., spatiotemporally optimal plans. Sixteen phantom cases with various sizes and locations of tumors and organs-at-risk (OAR) were generated using in-house software. Each case was planned with DART and conventional IMRT prescribing 60 Gy in 30 fractions. The observations of the change in the tumor volume over a treatment course were simulated using a two-level cell population model. Monte Carlo simulations of the treatment course for each case were run to account for uncertainty in the tumor response. The same OAR dose constraints were applied for both methods. The frequency of replanning was varied between 1, 2, 5 (weekly), and 29 times (daily). The final average tumor dose and OAR doses have been compared to quantify the potential dosimetric benefits of DART. RESULTS: The average tumor max, min, mean, and D95 doses using DART relative to these using conventional IMRT were 124.0%-125.2%, 102.1%-114.7%, 113.7%-123.4%, and 102.0%-115.9% (range dependent on the frequency of replanning). The average relative maximum doses for the cord and esophagus, mean doses for the heart and lungs, and D05 for the unspecified tissue resulting 84%-102.4%, 99.8%-106.9%, 66.9%-85.6%, 58.2%-78.8%, and 85.2%-94.0%, respectively. CONCLUSIONS: It is feasible to apply DART to the treatment of NSCLC using CBCT to observe the midtreatment tumor state. Potential increases in the tumor dose and reductions in the OAR dose, particularly for parallel OARs with mean or dose-volume constraints, could be achieved using DART compared to nonadaptive IMRT.

Personalized treatment planning with a model of radiation therapy outcomes for use in multiobjective optimization of IMRT plans for prostate cancer.

PURPOSE: To build a new treatment planning approach that extends beyond radiation transport and IMRT optimization by modeling the radiation therapy process and prognostic indicators for more outcome-focused decision making. METHODS: An in-house treatment planning system was modified to include multiobjective inverse planning, a probabilistic outcome model, and a multi-attribute decision aid. A genetic algorithm generated a set of plans embodying trade-offs between the separate objectives. An influence diagram network modeled the radiation therapy process of prostate cancer using expert opinion, results of clinical trials, and published research. A Markov model calculated a quality adjusted life expectancy (QALE), which was the endpoint for ranking plans. RESULTS: The Multiobjective Evolutionary Algorithm (MOEA) was designed to produce an approximation of the Pareto Front representing optimal tradeoffs for IMRT plans. Prognostic information from the dosimetrics of the plans, and from patient-specific clinical variables were combined by the influence diagram. QALEs were calculated for each plan for each set of patient characteristics. Sensitivity analyses were conducted to explore changes in outcomes for variations in patient characteristics and dosimetric variables. The model calculated life expectancies that were in agreement with an independent clinical study. CONCLUSIONS: The radiation therapy model proposed has integrated a number of different physical, biological and clinical models into a more comprehensive model. It illustrates a number of the critical aspects of treatment planning that can be improved and represents a more detailed description of the therapy process. A Markov model was implemented to provide a stronger connection between dosimetric variables and clinical outcomes and could provide a practical, quantitative method for making difficult clinical decisions.

A feasibility study: Selection of a personalized radiotherapy fractionation schedule using spatiotemporal optimization.

PURPOSE: To investigate the impact of using spatiotemporal optimization, i.e., intensity-modulated spatial optimization followed by fractionation schedule optimization, to select the patient-specific fractionation schedule that maximizes the tumor biologically equivalent dose (BED) under dose constraints for multiple organs-at-risk (OARs). METHODS: Spatiotemporal optimization was applied to a variety of lung tumors in a phantom geometry using a range of tumor sizes and locations. The optimal fractionation schedule for a patient using the linear-quadratic cell survival model depends on the tumor and OAR sensitivity to fraction size (α/ß), the effective tumor doubling time (Td), and the size and location of tumor target relative to one or more OARs (dose distribution). The authors used a spatiotemporal optimization method to identify the optimal number of fractions N that maximizes the 3D tumor BED distribution for 16 lung phantom cases. The selection of the optimal fractionation schedule used equivalent (30-fraction) OAR constraints for the heart (Dmean≤45 Gy), lungs (Dmean≤20 Gy), cord (Dmax≤45 Gy), esophagus (Dmax≤63 Gy), and unspecified tissues (D05≤60 Gy). To assess plan quality, the authors compared the minimum, mean, maximum, and D95 of tumor BED, as well as the equivalent uniform dose (EUD) for optimized plans to conventional intensity-modulated radiation therapy plans prescribing 60 Gy in 30 fractions. A sensitivity analysis was performed to assess the effects of Td (3-100 days), tumor lag-time (Tk=0-10 days), and the size of tumors on optimal fractionation schedule. RESULTS: Using an α/ß ratio of 10 Gy, the average values of tumor max, min, mean BED, and D95 were up to 19%, 21%, 20%, and 19% larger than those from conventional prescription, depending on Td and Tk used. Tumor EUD was up to 17% larger than the conventional prescription. For fast proliferating tumors with Td less than 10 days, there was no significant increase in tumor BED but the treatment course could be shortened without a loss in tumor BED. The improvement in the tumor mean BED was more pronounced with smaller tumors (p-value=0.08). CONCLUSIONS: Spatiotemporal optimization of patient plans has the potential to significantly improve local tumor control (larger BED/EUD) of patients with a favorable geometry, such as smaller tumors with larger distances between the tumor target and nearby OAR. In patients with a less favorable geometry and for fast growing tumors, plans optimized using spatiotemporal optimization and conventional (spatial-only) optimization are equivalent (negligible differences in tumor BED/EUD). However, spatiotemporal optimization yields shorter treatment courses than conventional spatial-only optimization. Personalized, spatiotemporal optimization of treatment schedules can increase patient convenience and help with the efficient allocation of clinical resources. Spatiotemporal optimization can also help identify a subset of patients that might benefit from nonconventional (large dose per fraction) treatments that are ineligible for the current practice of stereotactic body radiation therapy.

Bayesian network models for error detection in radiotherapy plans.

The purpose of this study is to design and develop a probabilistic network for detecting errors in radiotherapy plans for use at the time of initial plan verification. Our group has initiated a multi-pronged approach to reduce these errors. We report on our development of Bayesian models of radiotherapy plans. Bayesian networks consist of joint probability distributions that define the probability of one event, given some set of other known information. Using the networks, we find the probability of obtaining certain radiotherapy parameters, given a set of initial clinical information. A low probability in a propagated network then corresponds to potential errors to be flagged for investigation. To build our networks we first interviewed medical physicists and other domain experts to identify the relevant radiotherapy concepts and their associated interdependencies and to construct a network topology. Next, to populate the network's conditional probability tables, we used the Hugin Expert software to learn parameter distributions from a subset of de-identified data derived from a radiation oncology based clinical information database system. These data represent 4990 unique prescription cases over a 5 year period. Under test case scenarios with approximately 1.5% introduced error rates, network performance produced areas under the ROC curve of 0.88, 0.98, and 0.89 for the lung, brain and female breast cancer error detection networks, respectively. Comparison of the brain network to human experts performance (AUC of 0.90 ± 0.01) shows the Bayes network model performs better than domain experts under the same test conditions. Our results demonstrate the feasibility and effectiveness of comprehensive probabilistic models as part of decision support systems for improved detection of errors in initial radiotherapy plan verification procedures.

Evaluation of the new cesium-131 seed for use in low-energy x-ray brachytherapy.

Characterization measurements and calculations were performed on a new medical seed developed by IsoRay Inc. in Richland, Washington, that utilizes the short-lived isotope 131Cs. This model has recently received FDA 510(k) clearance. The objective of this work was to characterize the dosimetric properties of the new seed according to the AAPM Task Group 43 recommendations. Cesium-131 is a low-energy x-ray emitter, with the most prominent peaks in the 29 keV to 34 keV region. The intended application is brachytherapy for treating cancers in prostate, breast, head and neck, lung, and pancreas. The evaluations performed included air-kerma strength, radial dose function, anisotropy in phantom, half-life, energy spectra, and internal activity. The results indicate the CS-1 seeds have a dose-rate constant of 0.915 cGy hr(-1) U(-1) in water, dose penetration characteristics similar to 125I and 103Pd, anisotropy function values on the order of 0.71 at short distances and small angles, and an average anisotropy factor of 0.964. The overall dosimetric characteristics are similar to 125I and 103Pd seeds with the exception of half-life, which is 9.7 days, as compared to 17 days for 103Pd and 60 days for 125I. The shorter half-life may offer significant advantages in biological effectiveness.

Random and systematic beam modulator errors in dynamic intensity modulated radiotherapy.

This paper reports on the dosimetric effects of random and systematic modulator errors in delivery of dynamic intensity modulated beams. A sliding-widow type delivery that utilizes a combination of multileaf collimators (MLCs) and backup diaphragms was examined. Gaussian functions with standard deviations ranging from 0.5 to 1.5 mm were used to simulate random positioning errors. A clinical example involving a clival meningioma was chosen with optic chiasm and brain stem as limiting critical structures in the vicinity of the tumour. Dose calculations for different modulator fluctuations were performed, and a quantitative analysis was carried out based on cumulative and differential dose volume histograms for the gross target volume and surrounding critical structures. The study indicated that random modulator errors have a strong tendency to reduce minimum target dose and homogeneity. Furthermore, it was shown that random perturbation of both MLCs and backup diaphragms in the order of sigma = 1 mm can lead to 5% errors in prescribed dose. In comparison, when MLCs or backup diaphragms alone was perturbed, the system was more robust and modulator errors of at least sigma = 1.5 mm were required to cause dose discrepancies greater than 5%. For systematic perturbation, even errors in the order of +/- 0.5 mm were shown to result in significant dosimetric deviations.

Application of influence diagrams to prostate intensity-modulated radiation therapy plan selection.

The purpose is to incorporate clinically relevant factors such as patient-specific and dosimetric information as well as data from clinical trials in the decision-making process for the selection of prostate intensity-modulated radiation therapy (IMRT) plans. The approach is to incorporate the decision theoretic concept of an influence diagram into the solution of the multiobjective optimization inverse planning problem. A set of candidate IMRT plans was obtained by varying the importance factors for the planning target volume (PTV) and the organ-at-risk (OAR) in combination with simulated annealing to explore a large part of the solution space. The Pareto set for the PTV and OAR was analysed to demonstrate how the selection of the weighting factors influenced which part of the solution space was explored. An influence diagram based on a Bayesian network with 18 nodes was designed to model the decision process for plan selection. The model possessed nodes for clinical laboratory results, tumour grading, staging information, patient-specific information, dosimetric information, complications and survival statistics from clinical studies. A utility node was utilized for the decision-making process. The influence diagram successfully ranked the plans based on the available information. Sensitivity analyses were used to judge the reasonableness of the diagram and the results. In conclusion, influence diagrams lend themselves well to modelling the decision processes for IMRT plan selection. They provide an excellent means to incorporate the probabilistic nature of data and beliefs into one model. They also provide a means for introducing evidence-based medicine, in the form of results of clinical trials, into the decision-making process.