Search

PurposeThe aim was to develop a novel artificial intelligence (AI)-guided clinical decision support system, to predict radiation doses to subsites of the mandible using diagnostic computed tomography scans acquired before any planning of head and neck radiation therapy (RT).Methods and materialsA dose classifier was trained using RT plans from 86 patients with oropharyngeal cancer; the test set consisted of an additional 20 plans. The classifier was trained to predict whether mandible subsites would receive a mean dose >50 Gy. The AI predictions were prospectively evaluated and compared with those of a specialist head and neck radiation oncologist for 9 patients. Positive predictive value (PPV), negative predictive value (NPV), Pearson correlation coefficient, and Lin concordance correlation coefficient were calculated to compare the AI predictions to those of the physician.ResultsIn the test data set, the AI predictions had a PPV of 0.95 and NPV of 0.88. For 9 patients evaluated prospectively, there was a strong correlation between the predictions of the AI algorithm and physician (P = .72, P < .001). Comparing the AI algorithm versus the physician, the PPVs were 0.82 versus 0.25, and the NPVs were 0.94 versus 1.0, respectively. Concordance between physician estimates and final planned doses was 0.62; this was 0.71 between AI-based estimates and final planned doses.ConclusionAI-guided decision support increased precision and accuracy of pre-RT dental dose estimates.

PurposeDosimetric and clinical predictors of radiation-induced optic nerve/chiasm neuropathy (RION) after single-fraction stereotactic radiosurgery (SRS) or hypofractionated (2-5 fractions) radiosurgery (fSRS) were analyzed from pooled data that were extracted from published reports (PubMed indexed from 1990 to June 2015). This study was undertaken as part of the American Association of Physicists in Medicine Working Group on Stereotactic Body Radiotherapy, investigating normal tissue complication probability (NTCP) after hypofractionated radiation.Methods and materialsEligible studies described dose delivered to optic nerve/chiasm and provided crude or actuarial toxicity risks, with visual endpoints (ie, loss of visual acuity, alterations in visual fields, and/or blindness/complete vision loss). Studies of patients with optic nerve sheath tumors, optic nerve gliomas, or ocular/uveal melanoma were excluded to obviate direct tumor effects on visual outcomes, as were studies not specifying causes of vision loss (ie, tumor progression vs RION).ResultsThirty-four studies (1578 patients) were analyzed. Histologies included pituitary adenoma, cavernous sinus meningioma, craniopharyngioma, and malignant skull base tumors. Prior resection (76% of patients) did not correlate with RION risk (P = .66). Prior irradiation (6% of patients) was associated with a crude 10-fold increased RION risk versus no prior radiation therapy. In patients with no prior radiation therapy receiving SRS/fSRS in 1-5 fractions, optic apparatus maximum point doses resulting in

Deep learning algorithms have recently been developed that utilize patient anatomy and raw imaging information to predict radiation dose, as a means to increase treatment planning efficiency and improve radiotherapy plan quality. Current state-of-the-art techniques rely on convolutional neural networks (CNNs) that use pixel-to-pixel loss to update network parameters. However, stereotactic body radiotherapy (SBRT) dose is often heterogeneous, making it difficult to model using pixel-level loss. Generative adversarial networks (GANs) utilize adversarial learning that incorporates image-level loss and is better suited to learn from heterogeneous labels. However, GANs are difficult to train and rely on compromised architectures to facilitate convergence. This study suggests an attention-gated generative adversarial network (DoseGAN) to improve learning, increase model complexity, and reduce network redundancy by focusing on relevant anatomy. DoseGAN was compared to alternative state-of-the-art dose prediction algorithms using heterogeneity index, conformity index, and various dosimetric parameters. All algorithms were trained, validated, and tested using 141 prostate SBRT patients. DoseGAN was able to predict more realistic volumetric dosimetry compared to all other algorithms and achieved statistically significant improvement compared to all alternative algorithms for the V100 and V120 of the PTV, V60 of the rectum, and heterogeneity index.

Machine learning is proving invaluable across disciplines. However, its success is often limited by the quality and quantity of available data, while its adoption is limited by the level of trust afforded by given models. Human vs. machine performance is commonly compared empirically to decide whether a certain task should be performed by a computer or an expert. In reality, the optimal learning strategy may involve combining the complementary strengths of humans and machines. Here, we present expert-augmented machine learning (EAML), an automated method that guides the extraction of expert knowledge and its integration into machine-learned models. We used a large dataset of intensive-care patient data to derive 126 decision rules that predict hospital mortality. Using an online platform, we asked 15 clinicians to assess the relative risk of the subpopulation defined by each rule compared to the total sample. We compared the clinician-assessed risk to the empirical risk and found that, while clinicians agreed with the data in most cases, there were notable exceptions where they overestimated or underestimated the true risk. Studying the rules with greatest disagreement, we identified problems with the training data, including one miscoded variable and one hidden confounder. Filtering the rules based on the extent of disagreement between clinician-assessed risk and empirical risk, we improved performance on out-of-sample data and were able to train with less data. EAML provides a platform for automated creation of problem-specific priors, which help build robust and dependable machine-learning models in critical applications.

PurposeTo suggest an attention-aware, cycle-consistent generative adversarial network (A-CycleGAN) enhanced with variational autoencoding (VAE) as a superior alternative to current state-of-the-art MR-to-CT image translation methods.Materials and methodsAn attention-gating mechanism is incorporated into a discriminator network to encourage a more parsimonious use of network parameters, whereas VAE enhancement enables deeper discrimination architectures without inhibiting model convergence. Findings from 60 patients with head, neck, and brain cancer were used to train and validate A-CycleGAN, and findings from 30 patients were used for the holdout test set and were used to report final evaluation metric results using mean absolute error (MAE) and peak signal-to-noise ratio (PSNR).ResultsA-CycleGAN achieved superior results compared with U-Net, a generative adversarial network (GAN), and a cycle-consistent GAN. The A-CycleGAN averages, 95% confidence intervals (CIs), and Wilcoxon signed-rank two-sided test statistics are shown for MAE (19.61 [95% CI: 18.83, 20.39], P = .0104), structure similarity index metric (0.778 [95% CI: 0.758, 0.798], P = .0495), and PSNR (62.35 [95% CI: 61.80, 62.90], P = .0571).ConclusionA-CycleGANs were a superior alternative to state-of-the-art MR-to-CT image translation methods.© RSNA, 2020.

PurposeTo assess stereotactic radiotherapy (SRT)/stereotactic body radiotherapy (SBRT) practices by polling clinics participating in multi-institutional clinical trials.MethodsThe NRG Oncology Medical Physics Subcommittee distributed a survey consisting of 23 questions, which covered general technologies, policies, and procedures used in the Radiation Oncology field for the delivery of SRT/SBRT (9 questions), and site-specific questions for brain SRT, lung SBRT, and prostate SBRT (14 questions). Surveys were distributed to 1,996 radiotherapy institutions included on the membership rosters of the five National Clinical Trials Network (NCTN) groups. Patient setup, motion management, target localization, prescriptions, and treatment delivery technique data were reported back by 568 institutions (28%).Results97.5% of respondents treat lung SBRT patients, 77.0% perform brain SRT, and 29.1% deliver prostate SBRT. 48.8% of clinics require a physicist present for every fraction of SBRT, 18.5% require a physicist present for the initial SBRT fraction only, and 14.9% require a physicist present for the entire first fraction, including set-up approval for all subsequent fractions. 55.3% require physician approval for all fractions, and 86.7% do not reposition without x-ray imaging. For brain SRT, most institutions (83.9%) use a planning target volume (PTV) margin of 2 mm or less. Lung SBRT PTV margins of 3 mm or more are used in 80.6% of clinics. Volumetric modulated arc therapy (VMAT) is the dominant delivery method in 62.8% of SRT treatments, 70.9% of lung SBRT, and 68.3% of prostate SBRT.ConclusionThis report characterizes SRT/SBRT practices in radiotherapy clinics participating in clinical trials. Data made available here allows the radiotherapy community to compare their practice with that of other clinics, determine what is achievable, and assess areas for improvement.

The expansion of machine learning to high-stakes application domains such as medicine, finance, and criminal justice, where making informed decisions requires clear understanding of the model, has increased the interest in interpretable machine learning. The widely used Classification and Regression Trees (CART) have played a major role in health sciences, due to their simple and intuitive explanation of predictions. Ensemble methods like gradient boosting can improve the accuracy of decision trees, but at the expense of the interpretability of the generated model. Additive models, such as those produced by gradient boosting, and full interaction models, such as CART, have been investigated largely in isolation. We show that these models exist along a spectrum, revealing previously unseen connections between these approaches. This paper introduces a rigorous formalization for the additive tree, an empirically validated learning technique for creating a single decision tree, and shows that this method can produce models equivalent to CART or gradient boosted stumps at the extremes by varying a single parameter. Although the additive tree is designed primarily to provide both the model interpretability and predictive performance needed for high-stakes applications like medicine, it also can produce decision trees represented by hybrid models between CART and boosted stumps that can outperform either of these approaches.

BackgroundWe investigated prognostic models based on clinical, radiologic, and radiomic feature to preoperatively identify meningiomas at risk for poor outcomes.MethodsRetrospective review was performed for 303 patients who underwent resection of 314 meningiomas (57% World Health Organization grade I, 35% grade II, and 8% grade III) at two independent institutions, which comprised primary and external datasets. For each patient in the primary dataset, 16 radiologic and 172 radiomic features were extracted from preoperative magnetic resonance images, and prognostic features for grade, local failure (LF) or overall survival (OS) were identified using the Kaplan-Meier method, log-rank tests and recursive partitioning analysis. Regressions and random forests were used to generate and test prognostic models, which were validated using the external dataset.ResultsMultivariate analysis revealed that apparent diffusion coefficient hypointensity (HR 5.56, 95% CI 2.01-16.7, P = .002) was associated with high grade meningioma, and low sphericity was associated both with increased LF (HR 2.0, 95% CI 1.1-3.5, P = .02) and worse OS (HR 2.94, 95% CI 1.47-5.56, P = .002). Both radiologic and radiomic predictors of adverse meningioma outcomes were significantly associated with molecular markers of aggressive meningioma biology, such as somatic mutation burden, DNA methylation status, and FOXM1 expression. Integrated prognostic models combining clinical, radiologic, and radiomic features demonstrated improved accuracy for meningioma grade, LF, and OS (area under the curve 0.78, 0.75, and 0.78, respectively) compared to models based on clinical features alone.ConclusionsPreoperative radiologic and radiomic features such as apparent diffusion coefficient and sphericity can predict tumor grade, LF, and OS in patients with meningioma.

Monte Carlo (MC)-based dose calculations are generally superior to analytical dose calculations (ADC) in modeling the dose distribution for proton pencil beam scanning (PBS) treatments. The purpose of this paper is to present a methodology for commissioning and validating an accurate MC code for PBS utilizing a parameterized source model, including an implementation of a range shifter, that can independently check the ADC in commercial treatment planning system (TPS) and fast Monte Carlo dose calculation in opensource platform (MCsquare). The source model parameters (including beam size, angular divergence and energy spread) and protons per MU were extracted and tuned at the nozzle exit by comparing Tool for Particle Simulation (TOPAS) simulations with a series of commissioning measurements using scintillation screen/CCD camera detector and ionization chambers. The range shifter was simulated as an independent object with geometric and material information. The MC calculation platform was validated through comprehensive measurements of single spots, field size factors (FSF) and three-dimensional dose distributions of spread-out Bragg peaks (SOBPs), both without and with the range shifter. Differences in field size factors and absolute output at various depths of SOBPs between measurement and simulation were within 2.2%, with and without a range shifter, indicating an accurate source model. TOPAS was also validated against anthropomorphic lung phantom measurements. Comparison of dose distributions and DVHs for representative liver and lung cases between independent MC and analytical dose calculations from a commercial TPS further highlights the limitations of the ADC in situations of highly heterogeneous geometries. The fast MC platform has been implemented within our clinical practice to provide additional independent dose validation/QA of the commercial ADC for patient plans. Using the independent MC, we can more efficiently commission ADC by reducing the amount of measured data required for low dose "halo" modeling, especially when a range shifter is employed.

Background and purposeChest wall toxicity is observed after stereotactic body radiation therapy (SBRT) for peripherally located lung tumors. We utilize machine learning algorithms to identify toxicity predictors to develop dose-volume constraints.Materials and methodsTwenty-five patient, tumor, and dosimetric features were recorded for 197 consecutive patients with Stage I NSCLC treated with SBRT, 11 of whom (5.6%) developed CTCAEv4 grade ≥2 chest wall pain. Decision tree modeling was used to determine chest wall syndrome (CWS) thresholds for individual features. Significant features were determined using independent multivariate methods. These methods incorporate out-of-bag estimation using Random forests (RF) and bootstrapping (100 iterations) using decision trees.ResultsUnivariate analysis identified rib dose to 1 cc

BackgroundMeningiomas are stratified according to tumor grade and extent of resection, often in isolation of other clinical variables. Here, we use machine learning (ML) to integrate demographic, clinical, radiographic and pathologic data to develop predictive models for meningioma outcomes.Methods and findingsWe developed a comprehensive database containing information from 235 patients who underwent surgery for 257 meningiomas at a single institution from 1990 to 2015. The median follow-up was 4.3 years, and resection specimens were re-evaluated according to current diagnostic criteria, revealing 128 WHO grade I, 104 grade II and 25 grade III meningiomas. A series of ML algorithms were trained and tuned by nested resampling to create models based on preoperative features, conventional postoperative features, or both. We compared different algorithms' accuracy as well as the unique insights they offered into the data. Machine learning models restricted to preoperative information, such as patient demographics and radiographic features, had similar accuracy for predicting local failure (AUC = 0.74) or overall survival (AUC = 0.68) as models based on meningioma grade and extent of resection (AUC = 0.73 and AUC = 0.72, respectively). Integrated models incorporating all available demographic, clinical, radiographic and pathologic data provided the most accurate estimates (AUC = 0.78 and AUC = 0.74, respectively). From these models, we developed decision trees and nomograms to estimate the risks of local failure or overall survival for meningioma patients.ConclusionsClinical information has been historically underutilized in the prediction of meningioma outcomes. Predictive models trained on preoperative clinical data perform comparably to conventional models trained on meningioma grade and extent of resection. Combination of all available information can help stratify meningioma patients more accurately.

Machine learning (ML) has the potential to revolutionize the field of radiation oncology, but there is much work to be done. In this article, we approach the radiotherapy process from a workflow perspective, identifying specific areas where a data-centric approach using ML could improve the quality and efficiency of patient care. We highlight areas where ML has already been used, and identify areas where we should invest additional resources. We believe that this article can serve as a guide for both clinicians and researchers to start discussing issues that must be addressed in a timely manner.

PurposeTo validate a machine learning approach to Virtual intensity-modulated radiation therapy (IMRT) quality assurance (QA) for accurately predicting gamma passing rates using different measurement approaches at different institutions.MethodsA Virtual IMRT QA framework was previously developed using a machine learning algorithm based on 498 IMRT plans, in which QA measurements were performed using diode-array detectors and a 3%local/3 mm with 10% threshold at Institution 1. An independent set of 139 IMRT measurements from a different institution, Institution 2, with QA data based on portal dosimetry using the same gamma index, was used to test the mathematical framework. Only pixels with ≥10% of the maximum calibrated units (CU) or dose were included in the comparison. Plans were characterized by 90 different complexity metrics. A weighted poison regression with Lasso regularization was trained to predict passing rates using the complexity metrics as input.ResultsThe methodology predicted passing rates within 3% accuracy for all composite plans measured using diode-array detectors at Institution 1, and within 3.5% for 120 of 139 plans using portal dosimetry measurements performed on a per-beam basis at Institution 2. The remaining measurements (19) had large areas of low CU, where portal dosimetry has a larger disagreement with the calculated dose and as such, the failure was expected. These beams need further modeling in the treatment planning system to correct the under-response in low-dose regions. Important features selected by Lasso to predict gamma passing rates were as follows: complete irradiated area outline (CIAO), jaw position, fraction of MLC leafs with gaps smaller than 20 or 5 mm, fraction of area receiving less than 50% of the total CU, fraction of the area receiving dose from penumbra, weighted average irregularity factor, and duty cycle.ConclusionsWe have demonstrated that Virtual IMRT QA can predict passing rates using different measurement techniques and across multiple institutions. Prediction of QA passing rates can have profound implications on the current IMRT process.

To develop a patient-specific 'big data' clinical decision tool to predict pneumonitis in stage I non-small cell lung cancer (NSCLC) patients after stereotactic body radiation therapy (SBRT). 61 features were recorded for 201 consecutive patients with stage I NSCLC treated with SBRT, in whom 8 (4.0%) developed radiation pneumonitis. Pneumonitis thresholds were found for each feature individually using decision stumps. The performance of three different algorithms (Decision Trees, Random Forests, RUSBoost) was evaluated. Learning curves were developed and the training error analyzed and compared to the testing error in order to evaluate the factors needed to obtain a cross-validated error smaller than 0.1. These included the addition of new features, increasing the complexity of the algorithm and enlarging the sample size and number of events. In the univariate analysis, the most important feature selected was the diffusion capacity of the lung for carbon monoxide (DLCO adj%). On multivariate analysis, the three most important features selected were the dose to 15 cc of the heart, dose to 4 cc of the trachea or bronchus, and race. Higher accuracy could be achieved if the RUSBoost algorithm was used with regularization. To predict radiation pneumonitis within an error smaller than 10%, we estimate that a sample size of 800 patients is required. Clinically relevant thresholds that put patients at risk of developing radiation pneumonitis were determined in a cohort of 201 stage I NSCLC patients treated with SBRT. The consistency of these thresholds can provide radiation oncologists with an estimate of their reliability and may inform treatment planning and patient counseling. The accuracy of the classification is limited by the number of patients in the study and not by the features gathered or the complexity of the algorithm.

The aim of this study is to compare two methodologies of prostate localization in a large cohort of patients. Daily prostate localization using B-mode ultrasound has been performed at the Nebraska Medical Center since 2000. More recently, a technology using electromagnetic transponders implanted within the prostate was introduced into our clinic (Calypso(R)). With each technology, patients were localized initially using skin marks. Localization error distributions were determined from offsets between the initial setup positions and those determined by ultrasound or Calypso. Ultrasound localization data was summarized from 16619 imaging sessions spanning 7 years; Calypso localization data consists of 1524 fractions in 41 prostate patients treated in the course of a clinical trial at five institutions and 640 localizations from the first 16 patients treated with our clinical system. Ultrasound and Calypso patients treated between March and September 2007 at the Nebraska Medical Center were analyzed and compared, allowing a single institutional comparison of the two technologies. In this group of patients, the isocenter determined by ultrasound-based localization is on average 5.3 mm posterior to that determined by Calypso, while the systematic and random errors and PTV margins calculated from the ultrasound localizations were 3 - 4 times smaller than those calculated from the Calypso localizations. Our study finds that there are systematic differences between Calypso and ultrasound for prostate localization.

Respiratory gated radiotherapy may allow reduction of the treatment margins, thus sparing healthy tissue and/or allowing dose escalation to the tumor. However, current commissioning and quality assurance of linear accelerators do not include evaluation of gated delivery. The purpose of this study is to test gated photon delivery of a Siemens ONCOR Avant-Garde linear accelerator. Dosimetric characteristics for gated and nongated delivery of 6-MV and 15-MV photons were compared for the range of doses, dose rates, and for several gating regimes. Dose profiles were also compared using Kodak EDR2 and X-Omat V films for 6-MV and 15-MV photons for several dose rates and gating regimes. Results showed that deviation is less than or equal to 0.6% for all dose levels evaluated with the exception of the lowest dose delivered at 25 MU at an unrealistically high gating frequency of 0.5 Hz. At 400 MU, dose profile deviations along the central axes in in-plane and cross-plane directions within 80% of the field size are below 0.7%. No unequivocally detectable dose profile deviation was observed for 50 MU. Based on the comparison with widely accepted standards for conventional delivery, our results indicate that this LINAC is well suited for gated delivery of nondynamic fields.

A patient specific quality assurance program has been developed to facilitate the clinical implementation of intensity modulated radiotherapy (IMRT) delivered using a micro-multileaf collimator. The methodology includes several dosimetric tasks that are performed prior to the treatment of each patient. Film dosimetry is performed for each individual field and for the multifield composite plan. Individual field measurements are performed at a depth of 5 cm in a water equivalent slab phantom; export of dose calculations from the treatment planning system is similarly specified. For the composite distribution, parameters from the patient plan are applied to an IMRT phantom, and film is exposed in an axial orientation. Distributions are compared with the aid of software developed for the specific tasks. The measured and calculated dose distributions can be superimposed and positioned graphically using move, rotate, and mirror tools, as well as by specifying isocenter coordinates and using fiducial marks. Horizontal and vertical profiles are available for analysis. Dose difference, distance-to-agreement, and gamma index, the minimum scaled multidimensional distance between a measurement and a calculation point determined in combined dose and physical distance space, are calculated along a specified isodose line and displayed. gamma provides an excellent measure of disagreement between measurement and calculation for complex intensity distributions. We specify 3% dose difference and 3 mm distance as our scaling acceptability criteria. Absolute dosimetry for each composite plan is performed using an ionization chamber. To date, excellent agreement between measurements and calculations has been observed.

We describe a method for independently verifying the dose distributions from pre- and post-implant brachytherapy source distributions. Monte Carlo calculations have been performed to characterize the three-dimensional dose distribution in water phantom from a low-energy brachytherapy source. The calculations are performed in a voxelized, Cartesian coordinate geometry and normalized based upon a separate Monte Carlo calculation for the seed specific air-kerma strength to produce an absolute dose grid with units of cGy hr(-1) x U(-1). The seed-specific, three-dimensional dose grid is stored as a text file for processing using a separate visual basic program. This program requires the coordinate positions of each seed in the pre- or post-plan and sums the kernel file for a three-dimensional composite dose distribution. A kernel matrix size of 81x81x81 with a voxel size of 1.0x1.0x1.0 mm3 was chosen as a compromise between calculation time, kernel size, and truncation of the stored dose distribution as a function of radial distance from the midpoint of the seed. Good agreement is achieved for a representative pre- and post-plan comparison versus a commercial implementation of the TG-43 brachytherapy dosimetry protocol.