Quantifying Regional Radiation-Induced Lung Injury in Patients Using Hyperpolarized 129Xe Gas Exchange Magnetic Resonance Imaging.

PURPOSE: Radiation-induced lung injury has been shown to alter regional ventilation and perfusion in the lung. However, changes in regional pulmonary gas exchange have not previously been measured. METHODS AND MATERIALS: Ten patients receiving conventional radiation therapy (RT) for lung cancer underwent pre-RT and 3-month post-RT magnetic resonance imaging (MRI) using an established hyperpolarized 129Xe gas exchange technique to map lung function. Four patients underwent an additional 8-month post-RT MRI. The MR signal from inhaled xenon was measured in the following 3 pulmonary compartments: the lung airspaces, the alveolar membrane tissue, and the pulmonary capillaries (interacting with red blood cells [RBCs]). Thoracic 1H MRI scans were acquired, and deformable registration was used to transfer 129Xe functional maps to the RT planning computed tomography scan. The RT-associated changes in ventilation, membrane uptake, and RBC transfer were computed as a function of regional lung dose (equivalent dose in 2-Gy fractions). Pearson correlations and t tests were used to determine statistical significance, and weighted sum of squares linear regression subsequently characterized the dose dependence of each functional component. The pulmonary function testing metrics of forced vital capacity and diffusing capacity for carbon monoxide were also acquired at each time point. RESULTS: Compared with pre-RT baseline, 3-month post-RT ventilation decreased by an average of -0.24 ± 0.05%/Gy (ρ = -0.88; P < .001), membrane uptake increased by 0.69 ± 0.14%/Gy (ρ = 0.94; P < .001), and RBC transfer decreased by -0.41 ± 0.06%/Gy (ρ = -0.92; P < .001). Membrane uptake maintained a strong positive correlation with regional dose at 8 months post-RT, demonstrating an increase of 0.73 ± 0.11%/Gy (ρ = 0.92; P = .006). Changes in membrane uptake and RBC transfer appeared greater in magnitude (%/Gy) for individuals with low heterogeneity in their baseline lung function. An increase in whole-lung membrane uptake showed moderate correlation with decreases in forced vital capacity (ρ = -0.50; P = .17) and diffusing capacity for carbon monoxide (ρ = -0.44; P = .23), with neither correlation reaching statistical significance. CONCLUSIONS: Hyperpolarized 129Xe MRI measured and quantified regional, RT-associated, dose-dependent changes in pulmonary gas exchange. This tool could enable future work to improve our understanding and management of radiation-induced lung injury.

Augmenting Quality Assurance Measures in Treatment Review with Machine Learning in Radiation Oncology.

Purpose: Pretreatment quality assurance (QA) of treatment plans often requires a high cognitive workload and considerable time expenditure. This study explores the use of machine learning to classify pretreatment chart check QA for a given radiation plan as difficult or less difficult, thereby alerting the physicists to increase scrutiny on difficult plans. Methods and Materials: Pretreatment QA data were collected for 973 cases between July 2018 and October 2020. The outcome variable, a degree of difficulty, was collected as a subjective rating by physicists who performed the pretreatment chart checks. Potential features were identified based on clinical relevance, contribution to plan complexity, and QA metrics. Five machine learning models were developed: support vector machine, random forest classifier, adaboost classifier, decision tree classifier, and neural network. These were incorporated into a voting classifier, where at least 2 algorithms needed to predict a case as difficult for it to be classified as such. Sensitivity analyses were conducted to evaluate feature importance. Results: The voting classifier achieved an overall accuracy of 77.4% on the test set, with 76.5% accuracy on difficult cases and 78.4% accuracy on less difficult cases. Sensitivity analysis showed features associated with plan complexity (number of fractions, dose per monitor unit, number of planning structures, and number of image sets) and clinical relevance (patient age) were sensitive across at least 3 algorithms. Conclusions: This approach can be used to equitably allocate plans to physicists rather than randomly allocate them, potentially improving pretreatment chart check effectiveness by reducing errors propagating downstream.

Idiopathic Pneumonitis Syndrome After Total Body Irradiation in Pediatric Patients Undergoing Myeloablative Hematopoietic Stem Cell Transplantation: A PENTEC Comprehensive Review.

PURPOSE: Pulmonary complications, especially idiopathic pneumonitis syndrome (IPS), are potentially life altering or fatal sequelae of hematopoietic cell transplantation (HCT). Total body irradiation (TBI) as part of the conditioning regimen has been implicated in IPS. A comprehensive PENTEC (Pediatric Normal Tissues in the Clinic) review was performed to increase our understanding of the role of TBI in the development of acute, noninfectious IPS. METHODS AND MATERIALS: A systematic literature search was conducted using the MEDLINE, PubMed, and Cochrane library databases for articles describing pulmonary toxicity in children treated with HCT. Data pertaining to TBI and pulmonary endpoints were extracted. Risk of IPS was analyzed in relation to patient age, TBI dose, fractionation, dose rate, lung shielding, timing, and type of transplant, with the goal to better understand factors associated with this complication in children undergoing HCT. A logistic regression model was developed using a subset of studies with comparable transplant regimens and sufficient TBI data. RESULTS: Six studies met criteria for modeling of the correlation of TBI parameters with IPS; all consisted of pediatric patients undergoing allogeneic HCT with a cyclophosphamide-based chemotherapy regimen. IPS was variably defined, but all studies that reported IPS were included in this analysis. The mean incidence of post-HCT IPS was 16% (range, 4%-41%). Mortality from IPS, when it occurred, was high (median, 50%; range, 45%-100%). Fractionated TBI prescription doses encompassed a narrow range of 9 to 14 Gy. Many differing TBI methods were reported, and there was an absence of 3-dimensional dose analysis of lung blocking techniques. Thus, a univariate correlation between IPS and total TBI dose, dose fractionation, dose rate, or TBI technique could not be made. However, a model, built from these studies based on prescribed dose using a normalized dose parameter of equivalent dose in 2-Gy fractions (EQD2), adjusted for dose rate, suggested correlation with the development of IPS (P = .0004). The model-predicted odds ratio for IPS was 24.3 Gy-1 (95% confidence interval, 7.0-84.3). Use of TBI lung dose metrics (eg, midlung point dose) could not be successfully modeled, potentially because of dosimetric uncertainties in the actual delivered volumetric lung dose and imperfections in our modeling process. CONCLUSIONS: This PENTEC report is a comprehensive review of IPS in pediatric patients receiving fractionated TBI regimens for allogenic HCT. IPS was not clearly associated with 1 single TBI factor. Modeling using dose-rate adjusted EQD2 showed a response with IPS for allogeneic HCT using a cyclophosphamide-based chemotherapy regimen. Therefore, this model suggests IPS mitigation strategies can focus on not just the dose and dose per fraction but also the dose rate used in TBI. More data are needed to confirm this model and to determine the influence of chemotherapy regimens and contribution from graft-versus-host disease. The presence of confounding variables (eg, systemic chemotherapies) that affect risk, the narrow range of fractionated TBI doses found in the literature, and limitations of other reported data (eg, lung point dose) may have prevented a more straightforward link between IPS and total dose from being observed.

Evaluation of the clinical impact of the differences between planned and delivered dose in prostate cancer radiotherapy based on CT-on-rails IGRT and patient-reported outcome scores.

PURPOSE: To estimate the clinical impact of differences between delivered and planned dose using dose metrics and normal tissue complication probability (NTCP) modeling. METHODS: Forty-six consecutive patients with prostate adenocarcinoma between 2010 and 2015 treated with intensity-modulated radiation therapy (IMRT) and who had undergone computed tomography on rails imaging were included. Delivered doses to bladder and rectum were estimated using a contour-based deformable image registration method. The bladder and rectum NTCP were calculated using dose-response parameters applied to planned and delivered dose distributions. Seven urinary and gastrointestinal symptoms were prospectively collected using the validated prostate cancer symptom indices patient reported outcome (PRO) at pre-treatment, weekly treatment, and post-treatment follow-up visits. Correlations between planned and delivered doses against PRO were evaluated in this study. RESULTS: Planned mean doses to bladder and rectum were 44.9 ± 13.6 Gy and 42.8 ± 7.3 Gy, while delivered doses were 46.1 ± 13.4 Gy and 41.3 ± 8.7 Gy, respectively. D10cc for rectum was 64.1 ± 7.6 Gy for planned and 60.1 ± 9.3 Gy for delivered doses. NTCP values of treatment plan were 22.3% ± 8.4% and 12.6% ± 5.9%, while those for delivered doses were 23.2% ± 8.4% and 9.9% ± 8.3% for bladder and rectum, respectively. Seven of 25 patients with follow-up data showed urinary complications (28%) and three had rectal complications (12%). Correlations of NTCP values of planned and delivered doses with PRO follow-up data were random for bladder and moderate for rectum (0.68 and 0.67, respectively). CONCLUSION: Sensitivity of bladder to clinical variations of dose accumulation indicates that an automated solution based on a DIR that considers inter-fractional organ deformation could recommend intervention. This is intended to achieve additional rectum sparing in cases that indicate higher than expected dose accumulation early during patient treatment in order to prevent acute severity of bowel symptoms.

Impact of Oral Cavity Dosimetry on Patient Reported Xerostomia and Dysgeusia in the Setting of Deintensified Chemoradiotherapy.

Purpose: To determine the relationship between mean oral cavity (OC) dose (treated as a singular organ at risk) to patient reported xerostomia and dysgeusia. In addition, we will examine the relationship between oral cavity substructure doses to patient reported xerostomia and dysgeusia. All patients were treated in the setting of deintensification (60 Gy). Methods and Materials: In the study, 184 and 177 prospectively enrolled patients for de-escalated chemoradiotherapy (CRT) for human papillomavirus (HPV)-positive oropharyngeal cancer submitted PROs at 6 and 12 months, respectively using Patient-Reported Outcomes version of the Common Terminology Criteria for Adverse Events questionnaire. Patient's OC consisting of the following substructures were segmented: oral tongue, base of tongue, floor of mouth, hard and soft palate, cheek mucosa, and upper and lower lip mucosa. Ordinal logistic regression (no/mild vs moderate vs severe/very severe symptoms) was used to compare organs at risk dosimetry to patient reported xerostomia and dysgeusia at 6 and 12 months. Multivariate ordinal logistic regression models were generated. Results: Mean dose to the contralateral parotid (P = .04), OC (P = .04), and baseline patient reported xerostomia (P = .009) were significantly associated with xerostomia severity at 6 months. Only baseline xerostomia (P = .02) and mean dose to the contralateral submandibular gland (P = .0001) were significantly associated with xerostomia severity at 12 months. The only significant factor related to dysgeusia at either time point was mean dose to the OC at 12 months (P = .009). On examining substructures, the mean dose to the floor of mouth was implicated for the dose relationship to 6-month xerostomia (P = .04), and the oral tongue was found to be implicated for the relationship for 12-month dysgeusia (P = .04). Conclusions: The mean dose to the OC was found to relate to xerostomia symptoms at 6 months post-CRT and dysgeusia symptoms at 12 months post-CRT. The mean dose to the floor of mouth and oral tongue appeared to drive this relationship for xerostomia and dysgeusia symptoms, respectively. This work suggests the floor of mouth and oral tongue should be prioritized during planning over the rest of the OC. The effect of OC dose relative to other salivary structures for xerostomia appeared to depend on time post-CRT.

Feature Engineering for Interpretable Machine Learning for Quality Assurance in Radiation Oncology.

Chart checking is a time intensive process with high cognitive workload for physicists. Previous studies have partially automated and standardized chart checking, but limited studies implement data-driven approaches to reduce cognitive workload for quality assurance processes. This study aims to evaluate feature selection methods to improve the interpretability and transparency of machine learning models in predicting the degree of difficulty for a pretreatment physics chart check. We compare chi-square, mutual information, feature importance thresholding, and greedy feature selection for four different classifiers. Random forest has the highest performance with SMOTE oversampling using mutual information for feature selection (accuracy 84.0%, AUC 87.0%, precision 80.0%, recall 80.0%). This study demonstrates that feature selection methods can improve model interpretability and transparency.

Prospective assessment of sparing the parotid ducts via MRI sialography for reducing patient reported xerostomia.

PURPOSE: To assess the impact of prospectively sparing the parotid ducts via MRI sialography on patient reported xerostomia for those receiving definitive radiotherapy (RT) for oropharyngeal squamous cell carcinoma. METHODS AND MATERIALS: Thirty-eight patients with oropharynx cancer to be treated with definitive RT underwent pre-treatment MRI sialograms to localize their parotid ducts. The parotid ducts were maximally spared during treatment planning. Patients reported symptoms (PRO-CTCAE and QLQ-H&N35) were collected at 6 and 12 months post-RT and compared to a historical cohort who underwent conventional parotid gland mean dose sparing. Regression models were generated using parotid and submandibular gland doses with and without incorporating the dose to the parotid ducts to determine the impact of parotid duct dose on patient reported xerostomia. RESULTS: At 6 months post-RT, 12/26 (46%) patients reported ≥moderate xerostomia when undergoing parotid ductal sparing compared to 43/61 (70%) in the historical cohort (p = 0.03). At 12 months post-RT, 8/22 (36%) patients reported ≥moderate xerostomia when undergoing parotid ductal sparing compared to 34/68(50%) in the historical cohort (p = 0.08). Using nested logistic regression models, the mean parotid duct dose was found to significantly relate to patient reported xerostomia severity at 6 months post-RT (p = 0.04) and trended towards statistical significance at 12 months post-RT (p = 0.09). At both 6 and 12 months post-RT, the addition of mean parotid duct dose significantly improved model fit (p < 0.05). CONCLUSIONS: MRI sialography guided parotid duct sparing appears to reduce the rates of patient-reported xerostomia. Further, logistic regression analysis found parotid duct dose to be significantly associated with patient reported xerostomia. A significant improvement in model fit was observed when adding mean parotid duct dose compared to models that only contain mean parotid gland dose and mean contralateral submandibular gland dose.

Clinical Use of A Priori Knowledge of Organ-At-Risk Sparing During Radiation Therapy Treatment for Oropharyngeal Cancer: Dosimetric and Patient Reported Outcome Improvements.

PURPOSE: This study aimed to prospectively assess dosimetric and clinical effects of treatment planners having a priori knowledge of the maximum achievable dose sparing for organs at risk (OARs) for patients with oropharynx cancer receiving intensity modulated radiation therapy (RT). METHODS AND MATERIALS: We examined patients with oropharynx cancer who were treated in prospective clinical trials between February 2012 and April 2019 at our institution. A tool generating estimates of maximum achievable dose sparing for OARs (feasibility dose-volume histogram [FDVH]) was used clinically starting July 2016. Patients were divided into 2 cohorts: Before (ie, baseline) and after (ie, FDVH-guided) FDVH. Doses received by various OARs were compared with those estimated to be achievable per FDVH, and that difference was defined as the excess of feasible dose (EFD). Patient-reported outcome (PRO) questionnaires were completed at 3, 6, and 12 months after treatment. The baseline and FDVH-guided cohorts were compared in terms of EFD, plan quality metrics, and post-RT PRO assessments. RESULTS: A total of 139 patients were included in the analysis (60 in the baseline cohort, 79 in the FDVH-guided cohort). The FDVH-guided cohort had lower EFD to the contralateral parotid by 4.1 Gy, the ipsilateral parotid by 10.6 Gy, the larynx by 4.3 Gy, the oral cavity by 1.5 Gy, and the contralateral submandibular gland by 0.4 Gy. Plan quality metrics were similar between the cohorts. Less variation of EFD was seen in the FDVH-guided cohort for the parotid glands and contralateral submandibular gland (P < .05). The average post-RT PROs were better in the FVHD cohort versus baseline (particularly at the 6-month timepoint for dry mouth frequency, sticky saliva, meal enjoyment, severity of pain, and Eating Assessment Tool 10 composite [swallowing]; P < .05). CONCLUSIONS: Use of FDVH was associated with improved and less variable OAR sparing for clinically delivered plans. FDVH-guided patients had improved PROs compared with baseline with a variety of outcomes significantly improved at 6 months after treatment.

Hyperpolarized 129Xe Magnetic Resonance Imaging for Functional Avoidance Treatment Planning in Thoracic Radiation Therapy: A Comparison of Ventilation- and Gas Exchange-Guided Treatment Plans.

PURPOSE: To present a methodology to use pulmonary gas exchange maps to guide functional avoidance treatment planning in radiation therapy (RT) and evaluate its efficacy compared with ventilation-guided treatment planning. METHODS AND MATERIALS: Before receiving conventional RT for non-small cell lung cancer, 11 patients underwent hyperpolarized 129Xe gas exchange magnetic resonance imaging to map the distribution of xenon in its gas phase (ventilation) and transiently bound to red blood cells in the alveolar capillaries (gas exchange). Both ventilation and gas exchange maps were independently used to guide development of new functional avoidance treatment plans for every patient, while adhering to institutional dose-volume constraints for normal tissues and target coverage. Furthermore, dose-volume histogram (DVH)-based reoptimizations of the clinical plan, with reductions in mean lung dose (MLD) equal to the functional avoidance plans, were created to serve as the control group. To evaluate each plan (regardless of type), gas exchange maps, representing end-to-end lung function, were used to calculate gas exchange-weighted MLD (fMLD), gas exchange-weighted volume receiving ≥20 Gy (fV20), and mean dose in the highest gas exchanging 33% and 50% volumes of lung (MLD-f33% and MLD-f50%). Using each clinically approved plan as a baseline, the reductions in functional metrics were compared for ventilation-optimization, gas exchange optimization, and DVH-based reoptimization. Statistical significance was determined using the Freidman test, with subsequent subdivision when indicated by P values less than .10 and post hoc testing with Wilcoxon signed rank tests to determine significant differences (P < .05). Toxicity modeling was performed using an established function-based model to estimate clinical significance of the results. RESULTS: Compared with DVH-based reoptimization of the clinically approved plans, gas exchange-guided functional avoidance planning more effectively reduced the gas exchange-weighted metrics fMLD (average ± SD, -78 ± 79 cGy, compared with -45 ± 34 cGy; P = .03), MLD-f33% (-135 ± 136 cGy, compared with -52 ± 47 cGy; P = .004), and MLD-f50% (-96 ± 95 cGy, compared with -47 ± 40 cGy; P = .01). Comparing the 2 functional planning types, Gas Exchange-Guided planning more effectively reduced MLD-f33% compared with ventilation-guided planning (-64 ± 95; P = .009). For some patients, Gas Exchange-Guided functional avoidance plans demonstrated clinically significant reductions in model-predicted toxicity, more so than the accompanying ventilation-guided plans and DVH-based reoptimizations. CONCLUSION: Gas Exchange-Guided planning effectively reduced dose to high gas exchanging regions of lung while maintaining clinically acceptable plan quality. In many patients, ventilation-guided planning incidentally reduced dose to higher gas exchange regions, to a lesser extent. This methodology enables future prospective trials to examine patient outcomes.

NTCP modeling and dose-volume correlations for acute xerostomia and dry eye after whole brain radiation.

BACKGROUND: Whole brain radiation (WBRT) may lead to acute xerostomia and dry eye from incidental parotid and lacrimal exposure, respectively. We performed a prospective observational study to assess the incidence/severity of this toxicity. We herein perform a secondary analysis relating parotid and lacrimal dosimetric parameters to normal tissue complication probability (NTCP) rates and associated models. METHODS: Patients received WBRT to 25-40 Gy in 10-20 fractions using 3D-conformal radiation therapy without prospective delineation of the parotids or lacrimals. Patients completed questionnaires at baseline and 1 month post-WBRT. Xerostomia was assessed using the University of Michigan xerostomia score (scored 0-100, toxicity defined as ≥ 20 pt increase) and xerostomia bother score (scored from 0 to 3, toxicity defined as ≥ 2 pt increase). Dry eye was assessed using the Subjective Evaluation of Symptom of Dryness (SESoD, scored from 0 to 4, toxicity defined as ≥ 2 pt increase). The clinical data were fitted by the Lyman-Kutcher-Burman (LKB) and Relative Seriality (RS) NTCP models. RESULTS: Of 55 evaluable patients, 19 (35%) had ≥ 20 point increase in xerostomia score, 11 (20%) had ≥ 2 point increase in xerostomia bother score, and 13 (24%) had ≥ 2 point increase in SESoD score. For xerostomia, parotid V10Gy-V20Gy correlated best with toxicity, with AUC 0.68 for xerostomia score and 0.69-0.71 for bother score. The values for the D50, m and n parameters of the LKB model were 22.3 Gy, 0.84 and 1.0 for xerostomia score and 28.4 Gy, 0.55 and 1.0 for bother score, respectively. The corresponding values for the D50, γ and s parameters of the RS model were 23.5 Gy, 0.28 and 0.0001 for xerostomia score and 32.0 Gy, 0.45 and 0.0001 for bother score, respectively. For dry eye, lacrimal V10Gy-V15Gy were found to correlate best with toxicity, with AUC values from 0.67 to 0.68. The parameter values of the LKB model were 53.5 Gy, 0.74 and 1.0, whereas of the RS model were 54.0 Gy, 0.37 and 0.0001, respectively. CONCLUSIONS: Xerostomia was most associated with parotid V10Gy-V20Gy, and dry eye with lacrimal V10Gy-V15Gy. NTCP models were successfully created for both toxicities and may help clinicians refine dosimetric goals and assess levels of risk in patients receiving palliative WBRT.

Evaluation of a commercial DIR platform for contour propagation in prostate cancer patients treated with IMRT/VMAT.

PURPOSE: To assess the performance and limitations of contour propagation with three commercial deformable image registration (DIR) algorithms using fractional scans of CT-on-rails (CTOR) and Cone Beam CT (CBCT) in image guided prostate therapy patients treated with IMRT/VMAT. METHODS: Twenty prostate cancer patients treated with IMRT/VMAT were selected for analysis. A total of 453 fractions across those patients were analyzed. Image data were imported into MIM (MIM Software, Inc., Cleveland, OH) and three DIR algorithms (DIR Profile, normalized intensity-based (NIB) and shadowed NIB DIR algorithms) were applied to deformably register each fraction with the planning CT. Manually drawn contours of bladder and rectum were utilized for comparison against the DIR propagated contours in each fraction. Four metrics were utilized in the evaluation of contour similarity, the Hausdorff Distance (HD), Mean Distance to Agreement (MDA), Dice Similarity Coefficient (DSC), and Jaccard indices. A subfactor analysis was performed per modality (CTOR vs. CBCT) and time (fraction). Point estimates and 95% confidence intervals were assessed via a Linear Mixed Effect model for the contour similarity metrics. RESULTS: No statistically significant differences were observed between the DIR Profile and NIB algorithms. However, statistically significant differences were observed between the shadowed NIB and NIB algorithms for some of the DIR evaluation metrics. The Hausdorff Distance calculation showed the NIB propagated contours vs. shadowed NIB propagated contours against the manual contours were 14.82 mm vs. 8.34 mm for bladder and 15.87 mm vs. 11 mm for rectum, respectively. Similarly, the Mean Distance to Agreement calculation comparing the NIB propagated contours vs. shadowed NIB propagated contours against the manual contours were 2.43 mm vs. 0.98 mm for bladder and 2.57 mm vs. 1.00 mm for rectum, respectively. The Dice Similarity Coefficients comparing the NIB propagated contours and shadowed NIB propagated contours against the manual contours were 0.844 against 0.936 for bladder and 0.772 against 0.907 for rectum, respectively. The Jaccard indices comparing the NIB propagated contours and shadowed NIB propagated contours against the manual contours were 0.749 against 0.884 for bladder and 0.637 against 0.831 for rectum, respectively. The shadowed NIB DIR, which showed the closest agreement with the manual contours performed significantly better than the DIR Profile in all the comparisons. The OAR with the greatest agreement varied substantially across patients and image guided radiation therapy (IGRT) modality. Intra-patient variability of contour metric evaluation was insignificant across all the DIR algorithms. Statistical significance at α = 0.05 was observed for manual vs. deformably propagated contours for bladder for all the metrics except Hausdorff Distance (P = 0.01 for MDA, P = 0.02 for DSC, P = 0.01 for Jaccard), whereas the corresponding values for rectum were: P = 0.03 for HD, P = 0.01 for MDA, P < 0.01 for DSC, P < 0.01 for Jaccard. The performance of the different metrics varied slightly across the fractions of each patient, which indicates that weekly contour propagation models provide a reasonable approximation of the daily contour propagation models. CONCLUSION: The high variance of Hausdorff Distance across all automated methods for bladder indicates widely variable agreement across fractions for all patients. Lower variance across all modalities, methods, and metrics were observed for rectum. The shadowed NIB propagated contours were substantially more similar to the manual contours than the DIR Profile or NIB contours for both the CTOR and CBCT imaging modalities. The relationship of each algorithm to similarity with manual contours is consistent across all observed metrics and organs. Screening of image guidance for substantial differences in bladder and rectal filling compared with the planning CT reference could aid in identifying fractions for which automated DIR would prove insufficient.

Using Artificial Intelligence to Improve the Quality and Safety of Radiation Therapy.

Within artificial intelligence, machine learning (ML) efforts in radiation oncology have augmented the transition from generalized to personalized treatment delivery. Although their impact on quality and safety of radiation therapy has been limited, they are increasingly being used throughout radiation therapy workflows. Various data-driven approaches have been used for outcome prediction, CT simulation, clinical decision support, knowledge-based planning, adaptive radiation therapy, plan validation, machine quality assurance, and process quality assurance; however, there are many challenges that need to be addressed with the creation and usage of ML algorithms as well as the interpretation and dissemination of findings. In this review, the authors present current applications of ML in radiation oncology quality and safety initiatives, discuss challenges faced by the radiation oncology community, and suggest future directions.

Task Group 174 Report: Utilization of [18 F]Fluorodeoxyglucose Positron Emission Tomography ([18 F]FDG-PET) in Radiation Therapy.

The use of positron emission tomography (PET) in radiation therapy (RT) is rapidly increasing in the areas of staging, segmentation, treatment planning, and response assessment. The most common radiotracer is 18 F-fluorodeoxyglucose ([18 F]FDG), a glucose analog with demonstrated efficacy in cancer diagnosis and staging. However, diagnosis and RT planning are different endeavors with unique requirements, and very little literature is available for guiding physicists and clinicians in the utilization of [18 F]FDG-PET in RT. The two goals of this report are to educate and provide recommendations. The report provides background and education on current PET imaging systems, PET tracers, intensity quantification, and current utilization in RT (staging, segmentation, image registration, treatment planning, and therapy response assessment). Recommendations are provided on acceptance testing, annual and monthly quality assurance, scanning protocols to ensure consistency between interpatient scans and intrapatient longitudinal scans, reporting of patient and scan parameters in literature, requirements for incorporation of [18 F]FDG-PET in treatment planning systems, and image registration. The recommendations provided here are minimum requirements and are not meant to cover all aspects of the use of [18 F]FDG-PET for RT.

Correlation of Regional Lung Ventilation and Gas Transfer to Red Blood Cells: Implications for Functional-Avoidance Radiation Therapy Planning.

PURPOSE: To investigate the degree to which lung ventilation and gas exchange are regionally correlated, using the emerging technology of hyperpolarized (HP)-129Xe magnetic resonance imaging (MRI). METHODS AND MATERIALS: Hyperpolarized-129Xe MRI studies were performed on 17 institutional review board-approved human subjects, including 13 healthy volunteers, 1 emphysema patient, and 3 non-small cell lung cancer patients imaged before and approximately 11 weeks after radiation therapy (RT). Subjects inhaled 1 L of HP-129Xe mixture, followed by the acquisition of interleaved ventilation and gas exchange images, from which maps were obtained of the relative HP-129Xe distribution in three states: (1) gaseous, in lung airspaces; (2) dissolved interstitially, in alveolar barrier tissue; and (3) transferred to red blood cells (RBCs), in the capillary vasculature. The relative spatial distributions of HP-129Xe in airspaces (regional ventilation) and RBCs (regional gas transfer) were compared. Further, we investigated the degree to which ventilation and RBC transfer images identified similar functional regions of interest (ROIs) suitable for functionally guided RT. For the RT patients, both ventilation and RBC functional images were used to calculate differences in the lung dose-function histogram and functional effective uniform dose. RESULTS: The correlation of ventilation and RBC transfer was ρ = 0.39 ± 0.15 in healthy volunteers. For the RT patients, this correlation was ρ = 0.53 ± 0.02 before treatment and ρ = 0.39 ± 0.07 after treatment; for the emphysema patient it was ρ = 0.24. Comparing functional ROIs, ventilation and RBC transfer demonstrated poor spatial agreement: Dice similarity coefficient = 0.50 ± 0.07 and 0.26 ± 0.12 for the highest-33%- and highest-10%-function ROIs in healthy volunteers, and in RT patients (before treatment) these were 0.58 ± 0.04 and 0.40 ± 0.04. The average magnitude of the differences between RBC- and ventilation-derived functional effective uniform dose, fV20Gy, fV10Gy, and fV5Gy were 1.5 ± 1.4 Gy, 4.1% ± 3.8%, 5.0% ± 3.8%, and 5.3% ± 3.9%, respectively. CONCLUSION: Ventilation may not be an effective surrogate for true regional lung function for all patients.

Fitting NTCP models to bladder doses and acute urinary symptoms during post-prostatectomy radiotherapy.

BACKGROUND: To estimate the radiobiological parameters of three popular normal tissue complication probability (NTCP) models, which describe the dose-response relations of bladder regarding different acute urinary symptoms during post-prostatectomy radiotherapy (RT). To evaluate the goodness-of-fit and the correlation of those models with those symptoms. METHODS: Ninety-three consecutive patients treated from 2010 to 2015 with post-prostatectomy image-guided intensity modulated radiotherapy (IMRT) were included in this study. Patient-reported urinary symptoms were collected pre-RT and weekly during treatment using the validated Prostate Cancer Symptom Indices (PCSI). The assessed symptoms were flow, dysuria, urgency, incontinence, frequency and nocturia using a Likert scale of 1 to 4 or 5. For this analysis, an increase by ≥2 levels in a symptom at any time during treatment compared to baseline was considered clinically significant. The dose volume histograms of the bladder were calculated. The Lyman-Kutcher-Burman (LKB), Relative Seriality (RS) and Logit NTCP models were used to fit the clinical data. The fitting of the different models was assessed through the area under the receiver operating characteristic curve (AUC), Akaike information criterion (AIC) and Odds Ratio methods. RESULTS: For the symptoms of urinary urgency, leakage, frequency and nocturia, the derived LKB model parameters were: 1) D50 = 64.2Gy, m = 0.50, n = 1.0; 2) D50 = 95.0Gy, m = 0.45, n = 0.50; 3) D50 = 83.1Gy, m = 0.56, n = 1.00; and 4) D50 = 85.4Gy, m = 0.60, n = 1.00, respectively. The AUC values for those symptoms were 0.66, 0.58, 0.64 and 0.64, respectively. The differences in AIC between the different models were less than 2 and ranged within 0.1 and 1.3. CONCLUSIONS: Different dose metrics were correlated with the symptoms of urgency, incontinence, frequency and nocturia. The symptoms of urinary flow and dysuria were poorly associated with dose. The values of the parameters of three NTCP models were determined for bladder regarding four acute urinary symptoms. All the models could fit the clinical data equally well. The NTCP predictions of urgency showed the best correlation with the patient reported outcomes.

Imaging Radiation-Induced Normal Tissue Injury to Quantify Regional Dose Response.

Noninvasive imaging has and will continue to play a pivotal role in the assessment of radiation-induced normal tissue toxicity. In this review, we will examine key literature regarding the use of anatomic and physiological imaging in relation to radiation-induced normal tissue toxicity. Additionally, this review contains a novel methodology for potentially incorporating dose-response data into treatment planning and normal tissue toxicity modeling.