Predictors Of Radiation Pneumonitis Research Articles

From pixels to prognosis: unveiling radiomics models with SHAP and LIME for enhanced interpretability

Radiomics-based prediction models have shown promise in predicting Radiation Pneumonitis (RP), a common adverse outcome of chest irradiation. Τhis study looks into more than just RP: it also investigates a bigger shift in the way radiomics-based models work. By integrating multi-modal radiomic data, which includes a wide range of variables collected from medical images including cutting-edge PET/CT imaging, we have developed predictive models that capture the intricate nature of illness progression. Radiomic features were extracted using PyRadiomics, encompassing intensity, texture, and shape measures. The high-dimensional dataset formed the basis for our predictive models, primarily Gradient Boosting Machines (GBM)—XGBoost, LightGBM, and CatBoost. Performance evaluation metrics, including Multi-Modal AUC-ROC, Sensitivity, Specificity, and F1-Score, underscore the superiority of the Deep Neural Network (DNN) model. The DNN achieved a remarkable Multi-Modal AUC-ROC of 0.90, indicating superior discriminatory power. Sensitivity and specificity values of 0.85 and 0.91, respectively, highlight its effectiveness in detecting positive occurrences while accurately identifying negatives. External validation datasets, comprising retrospective patient data and a heterogeneous patient population, validate the robustness and generalizability of our models. The focus of our study is the application of sophisticated model interpretability methods, namely SHAP (SHapley Additive exPlanations) and LIME (Local Interpretable Model-Agnostic Explanations), to improve the clarity and understanding of predictions. These methods allow clinicians to visualize the effects of features and provide localized explanations for every prediction, enhancing the comprehensibility of the model. This strengthens trust and collaboration between computational technologies and medical competence. The integration of data-driven analytics and medical domain expertise represents a significant shift in the profession, advancing us from analyzing pixel-level information to gaining valuable prognostic insights.

Synergizing the interaction of single nucleotide polymorphisms with dosiomics features to build a dual-omics model for the prediction of radiation pneumonitis

ObjectiveRadiation pneumonitis (RP) is the major dose-limiting toxicity of thoracic radiotherapy. This study aimed to developed a dual-omics (single nucleotide polymorphisms, SNP and dosiomics) prediction model for symptomatic RP. Materials and methodsThe potential SNPs, which are of significant difference between the RP grade ≥ 3 group and the RP grade ≤ 1 group, were selected from the whole exome sequencing SNPs using the Fisher’s exact test. Patients with lung cancer who received thoracic radiotherapy at our institution from 2009 to 2016 were enrolled for SNP selection and model construction. The factorization machine (FM) method was used to model the SNP epistasis effect, and to construct the RP prediction model (SNP-FM). The dosiomics features were extracted, and further selected using the minimum redundancy maximum relevance (mRMR) method. The selected dosiomics features were added to the SNP-FM model to construct the dual-omics model. ResultsFor SNP screening, peripheral blood samples of 28 patients with RP grade ≥ 3 and the matched 28 patients with RP grade ≤ 1 were sequenced. 81 SNPs were of significant difference (P < 0.015) and considered as potential SNPs. In addition, 21 radiation toxicity related SNPs were also included. For model construction, 400 eligible patients (including 108 RP grade ≥ 2) were enrolled. Single SNP showed no strong correlation with RP. On the other hand, the SNP-SNP interaction (epistasis effect) of 19 SNPs were modeled by the FM method, and achieved an area under the curve (AUC) of 0.76 in the testing group. In addition, 4 dosiomics features were selected and added to the model, and increased the AUC to 0.81. ConclusionsA novel dual-omics model by synergizing the SNP epistasis effect with dosiomics features was developed. The enhanced the RP prediction suggested its promising clinical utility in identifying the patients with severe RP during thoracic radiotherapy.

Enhancing the prediction of symptomatic radiation pneumonitis for locally advanced non-small-cell lung cancer by combining 3D deep learning-derived imaging features with dose-volume metrics: atwo-center study.

This study aims to examine the ability of deep learning (DL)-derived imaging features for the prediction of radiation pneumonitis (RP) in locally advanced non-small-cell lung cancer (LA-NSCLC) patients. The study cohort consisted of 90patients from the Fudan University Shanghai Cancer Center and 59patients from the Affiliated Hospital of Jiangnan University. Occurrences of RP were used as the endpoint event. Atotal of 512 3D DL-derived features were extracted from two regions of interest (lung-PTV and PTV-GTV) delineated on the pre-radiotherapy planning CT. Feature selection was done using LASSO regression, and the classification models were built using the multilayered perceptron method. Performances of the developed models were evaluated by receiver operating characteristic curve analysis. In addition, the developed models were supplemented with clinical variables and dose-volume metrics of relevance to search for increased predictive value. The predictive model using DL features derived from lung-PTV outperformed the one based on features extracted from PTV-GTV, with AUCs of 0.921 and 0.892, respectively, in the internal test dataset. Furthermore, incorporating the dose-volume metric V30Gy into the predictive model using features from lung-PTV resulted in an improvement of AUCs from 0.835 to 0.881 for the training data and from 0.690 to 0.746 for the validation data, respectively (DeLong p < 0.05). Imaging features extracted from pre-radiotherapy planning CT using 3D DL networks could predict radiation pneumonitis and may be of clinical value for risk stratification and toxicity management in LA-NSCLC patients. Integrating DL-derived features with dose-volume metrics provides apromising noninvasive method to predict radiation pneumonitis in LA-NSCLC lung cancer radiotherapy, thus improving individualized treatment and patient outcomes.

A dynamic nomogram predicting symptomatic pneumonia in patients with lung cancer receiving thoracic radiation

PurposeThe most common and potentially fatal side effect of thoracic radiation therapy is radiation pneumonitis (RP). Due to the lack of effective treatments, predicting radiation pneumonitis is crucial. This study aimed to develop a dynamic nomogram to accurately predict symptomatic pneumonitis (RP ≥ 2) following thoracic radiotherapy for lung cancer patients.MethodsData from patients with pathologically diagnosed lung cancer at the Zhongshan People’s Hospital Department of Radiotherapy for Thoracic Cancer between January 2017 and June 2022 were retrospectively analyzed. Risk factors for radiation pneumonitis were identified through multivariate logistic regression analysis and utilized to construct a dynamic nomogram. The predictive performance of the nomogram was validated using a bootstrapped concordance index and calibration plots.ResultsAge, smoking index, chemotherapy, and whole lung V5/MLD were identified as significant factors contributing to the accurate prediction of symptomatic pneumonitis. A dynamic nomogram for symptomatic pneumonitis was developed using these risk factors. The area under the curve was 0.89(95% confidence interval 0.83–0.95). The nomogram demonstrated a concordance index of 0.89(95% confidence interval 0.82–0.95) and was well calibrated. Furthermore, the threshold values for high- risk and low- risk were determined to be 154 using the receiver operating curve.ConclusionsThe developed dynamic nomogram offers an accurate and convenient tool for clinical application in predicting the risk of symptomatic pneumonitis in patients with lung cancer undergoing thoracic radiation.

Deep-Learning Model Prediction of Radiation Pneumonitis Using Pretreatment Chest Computed Tomography and Clinical Factors.

Objectives: This study aimed to build a comprehensive deep-learning model for the prediction of radiation pneumonitis using chest computed tomography (CT), clinical, dosimetric, and laboratory data. Introduction: Radiation therapy is an effective tool for treating patients with lung cancer. Despite its effectiveness, the risk of radiation pneumonitis limits its application. Although several studies have demonstrated models to predict radiation pneumonitis, no reliable model has been developed yet. Herein, we developed prediction models using pretreatment chest CT and various clinical data to assess the likelihood of radiation pneumonitis in lung cancer patients. Methods: This retrospective study analyzed 3-dimensional (3D) lung volume data from chest CT scans and 27 features including dosimetric, clinical, and laboratory data from 548 patients who were treated at our institution between 2010 and 2021. We developed a neural network, named MergeNet, which processes lung 3D CT, clinical, dosimetric, and laboratory data. The MergeNet integrates a convolutional neural network with subsequent fully connected layers. A support vector machine (SVM) and light gradient boosting machine (LGBM) model were also implemented for comparison. For comparison, the convolution-only neural network was implemented as well. Three-dimensional Resnet-10 network and 4-fold cross-validation were used. Results: Classification performance was quantified by using the area under the receiver operative characteristic curve (AUC) metrics. MergeNet showed the AUC of 0.689. SVM, LGBM, and convolution-only networks showed AUCs of 0.525, 0.541, and 0.550, respectively. Application of DeLong test to pairs of receiver operating characteristic curves respectively yielded P values of .001 for the MergeNet-SVM pair and 0.001 for the MergeNet-LGBM pair. Conclusion: The MergeNet model, which incorporates chest CT, clinical, dosimetric, and laboratory data, demonstrated superior performance compared to other models. However, since its prediction performance has not yet reached an efficient level for clinical application, further research is required. Contribution: This study showed that MergeNet may be an effective means to predict radiation pneumonitis. Various predictive factors can be used together for the radiation pneumonitis prediction task via the MergeNet.

Predicting radiation pneumonitis in lung cancer: a EUD-based machine learning approach for volumetric modulated arc therapy patients.

This study aims to develop an optimal machine learning model that uses lung equivalent uniform dose (lung EUD to predict radiation pneumonitis (RP) occurrence in lung cancer patients treated with volumetric modulated arc therapy (VMAT). We analyzed a cohort of 77 patients diagnosed with locally advanced squamous cell lung cancer (LASCLC) receiving concurrent chemoradiotherapy with VMAT. Patients were categorized based on the onset of grade II or higher radiation pneumonitis (RP 2+). Dose volume histogram data, extracted from the treatment planning system, were used to compute the lung EUD values for both groups using a specialized numerical analysis code. We identified the parameter α, representing the most significant relative difference in lung EUD between the two groups. The predictive potential of variables for RP2+, including physical dose metrics, lung EUD, normal tissue complication probability (NTCP) from the Lyman-Kutcher-Burman (LKB) model, and lung EUD-calibrated NTCP for affected and whole lung, underwent both univariate and multivariate analyses. Relevant variables were then employed as inputs for machine learning models: multiple logistic regression (MLR), support vector machine (SVM), decision tree (DT), and K-nearest neighbor (KNN). Each model's performance was gauged using the area under the curve (AUC), determining the best-performing model. The optimal α-value for lung EUD was 0.3, maximizing the relative lung EUD difference between the RP 2+ and non-RP 2+ groups. A strong correlation coefficient of 0.929 (P< 0.01) was observed between lung EUD (α = 0.3) and physical dose metrics. When examining predictive capabilities, lung EUD-based NTCP for the affected lung (AUC: 0.862) and whole lung (AUC: 0.815) surpassed LKB-based NTCP for the respective lungs. The decision tree (DT) model using lung EUD-based predictors emerged as the superior model, achieving an AUC of 0.98 in both training and validation datasets. The likelihood of developing RP 2+ has shown a significant correlation with the advancements in RT technology. From traditional 3-D conformal RT, lung cancer treatment methodologies have transitioned to sophisticated techniques like static IMRT. Accurately deriving such a dose-effect relationship through NTCP modeling of RP incidence is statistically challenging due to the increased number of degrees-of-freedom. To the best of our knowledge, many studies have not clarified the rationale behind setting the α-value to 0.99 or 1, despite the closely aligned calculated lung EUD and lung mean dose MLD. Perfect independence among variables is rarely achievable in real-world scenarios. Four prominent machine learning algorithms were used to devise our prediction models. The inclusion of lung EUD-based factors substantially enhanced their predictive performance for RP 2+. Our results advocate for the decision tree model with lung EUD-based predictors as the optimal prediction tool for VMAT-treated lung cancer patients. Which could replace conventional dosimetric parameters, potentially simplifying complex neural network structures in prediction models.

Inhibition of TGFβ1 activation prevents radiation-induced lung fibrosis.

Radiotherapy is the main treatment modality for thoracic tumours, but it may induce pulmonary fibrosis. Currently, the pathogenesis of radiation-induced pulmonary fibrosis (RIPF) is unclear, and effective treatments are lacking. Transforming growth factor beta 1 (TGFβ1) plays a central role in RIPF. We found that activated TGFβ1 had better performance for radiation pneumonitis (RP) risk prediction by detecting activated and total TGFβ1 levels in patient serum. αv integrin plays key roles in TGFβ1 activation, but the role of αv integrin-mediated TGFβ1 activation in RIPF is unclear. Here, we investigated the role of αv integrin-mediated TGFβ1 activation in RIPF and the application of the integrin antagonist cilengitide to prevent RIPF. ItgavloxP/loxP ;Pdgfrb-Cre mice were generated by conditionally knocking out Itgav in myofibroblasts, and wild-type mice were treated with cilengitide or placebo. All mice received 16Gy of radiation or underwent a sham radiation procedure. Lung fibrosis was measured by a modified Ashcroft score and microcomputed tomography (CT). An enzyme-linked immunosorbent assay (ELISA) was used to measure the serum TGFβ1 concentration, and total Smad2/3 and p-Smad2/3 levels were determined via Western blotting. Conditional Itgav knockout significantly attenuated RIPF (p<.01). Hounsfield units (HUs) in the lungs were reduced in the knockout mice compared with the control mice (p<.001). Conditional Itgav knockout decreased active TGFβ1 secretion and inhibited fibroblast p-Smad2/3 expression. Exogenous active TGFβ1, but not latent TGFβ1, reversed these reductions. Furthermore, cilengitide treatment elicited similar results and prevented RIPF. The present study revealed that conditional Itgav knockout and cilengitide treatment both significantly attenuated RIPF in mice by inhibiting αv integrin-mediated TGFβ1 activation. Activated TGFβ1 has a superior capacity in predicting radiation pneumonitis (RP) risk and plays a vital role in the development of radiation-induced pulmonary fibrosis (RIPF). Conditional knock out Itgav in myofibroblasts prevented mice from developing RIPF. Cilengitide alleviated the development of RIPF by inhibiting αv integrin-mediated TGFβ1 activation and may be used in targeted approaches for preventing RIPF.

Prediction of radiation pneumonitis using the effective α/β of lungs and heart in NSCLC patients treated with proton beam therapy

PurposeRadiation pneumonitis (RP) remains a major complication in non-small cell lung cancer (NSCLC) patients undergoing radiochemotherapy (RCHT). Traditionally, the mean lung dose (MLD) and the volume of the total lung receiving at least 20 Gy (V20Gy) are used to predict RP in patients treated with normo-fractionated photon therapy.However, other models, including the actual dose-distribution in the lungs using the effective α/β model or a combination of radiation doses to the lungs and heart, have been proposed for predicting RP. Moreover, the models established for photons may not hold for patients treated with passively-scattered proton therapy (PSPT). Therefore, we here tested and validated novel predictive parameters for RP in NSCLC patient treated with PSPT. MethodsData on the occurrence of RP, structure files and dose-volume histogram parameters for lungs and heart of 96 NSCLC patients, treated with PSPT and concurrent chemotherapy, was retrospectively retrieved from prospective clinical studies of two international centers. Data was randomly split into a training set (64 patients) and a validation set (32 patients). Statistical analyses were performed using binomial logistic regression. ResultsThe biologically effective dose (BED) of the’lungs - GTV’ significantly predicted RP ≥ grade 2 in the training-set using both a univariate model (p = 0.019, AUCtrain = 0.72) and a multivariate model in combination with the effective α/β parameter of the heart (pBED = 0.006, pα/βeff = 0.043, AUCtrain = 0.74). However, these results did not hold in the validation-set (AUCval = 0.52 andAUCval = 0.50, respectively). Moreover, these models were found to neither outperform a model built with the MLD (p = 0.015, AUCtrain = 0.73, AUCval = 0.51), nor a multivariate model additionally including the V20Gy of the heart (pMLD = 0.039, pV20Gy,heart = 0.58, AUCtrain = 0.74, AUCval = 0.53). ConclusionUsing the effective α/β parameter of the lungs and heart we achieved similar performance to commonly used models built for photon therapy, such as MLD, in predicting RP ≥ grade 2. Therefore, prediction models developed for photon RCHT still hold for patients treated with PSPT.

Quantification of diffuse parenchymal lung disease in non-small cell lung cancer patients with definitive concurrent chemoradiation therapy for predicting radiation pneumonitis.

We sought to quantify diffuse parenchymal lung disease (DPLD) extent using quantitative computed tomography (CT) analysis and to investigate its association with radiation pneumonitis (RP) development in non-small cell lung cancer (NSCLC) patients receiving definitive concurrent chemoradiation therapy (CCRT). A total of 82 NSCLC patients undergoing definitive CCRT were included in this prospective cohort study. Pretreatment CT scans were analyzed using quantitative CT analysis software. Low-attenuation area (LAA) features based on lung density and texture features reflecting interstitial lung disease (ILD) were extracted from the whole lung. Clinical and dosimetric factors were also evaluated. RP development was assessed using the Common Terminology Criteria for Adverse Events version 5.0. Univariable and multivariable logistic regression analyses were performed to identify independent risk factors for grade ≥3 (≥GR3) RP. RP was identified in 68 patients (73.9%), with nine patients (10.9%) experiencing ≥GR3 RP. Univariable logistic regression analysis identified excess kurtosis and high-attenuation area (HAA)_volume (cc) as significantly associated with ≥GR3 RP. Multivariable logistic regression analysis showed that the combined use of imaging features and clinical factors (forced expiratory volume in 1 second [FEV1], forced vital capacity [FVC], and CHEMO regimen) demonstrated the best performance (area under the receiver operating characteristic curve = 0.924) in predicting ≥GR3 RP. Quantified imaging features of DPLD obtained from pretreatment CT scans would predict the occurrence of RP in NSCLC patients undergoing definitive CCRT. Combining imaging features with clinical factors could improve the accuracy of the predictive model for severe RP.

Radiation pneumonia predictive model for radiotherapy in esophageal carcinoma patients

BackgroundThe machine learning models with dose factors and the deep learning models with dose distribution matrix have been used to building lung toxics models for radiotherapy and achieve promising results. However, few studies have integrated clinical features into deep learning models. This study aimed to explore the role of three-dimension dose distribution and clinical features in predicting radiation pneumonitis (RP) in esophageal cancer patients after radiotherapy and designed a new hybrid deep learning network to predict the incidence of RP.MethodsA total of 105 esophageal cancer patients previously treated with radiotherapy were enrolled in this study. The three-dimension (3D) dose distributions within the lung were extracted from the treatment planning system, converted into 3D matrixes and used as inputs to predict RP with ResNet. In total, 15 clinical factors were normalized and converted into one-dimension (1D) matrixes. A new prediction model (HybridNet) was then built based on a hybrid deep learning network, which combined 3D ResNet18 and 1D convolution layers. Machine learning-based prediction models, which use the traditional dosiomic factors with and without the clinical factors as inputs, were also constructed and their predictive performance compared with that of HybridNet using tenfold cross validation. Accuracy and area under the receiver operator characteristic curve (AUC) were used to evaluate the model effect. DeLong test was used to compare the prediction results of the models.ResultsThe deep learning-based model achieved superior prediction results compared with machine learning-based models. ResNet performed best in the group that only considered dose factors (accuracy, 0.78 ± 0.05; AUC, 0.82 ± 0.25), whereas HybridNet performed best in the group that considered both dose factors and clinical factors (accuracy, 0.85 ± 0.13; AUC, 0.91 ± 0.09). HybridNet had higher accuracy than that of Resnet (p = 0.009).ConclusionBased on prediction results, the proposed HybridNet model could predict RP in esophageal cancer patients after radiotherapy with significantly higher accuracy, suggesting its potential as a useful tool for clinical decision-making. This study demonstrated that the information in dose distribution is worth further exploration, and combining multiple types of features contributes to predict radiotherapy response.

The relationship of body mass index to setup errors, dosimetric parameters and incidence of radiation pneumonitis in non-small cell lung cancer patients undergoing intensity-modulated radiation therapy: a single-center observational study

Background The relationship among body mass index (BMI), setup error and radiation pneumonitis is not clearly illustrated. Objective The present study aimed to investigate the role of BMI in non-small cell lung cancer (NSCLC) patients’ radiation treatment, focusing on its relationship with setup error of patient positioning, the dosimetric parameters of intensity-modulated radiation therapy (IMRT) and the incidence of radiation pneumonitis. Methods This prospective observational study included 523 cases of NSCLC patients during 2020–2022. Patients were divided into different groups by different BMI. The setup error was obtained by cone beam CT (CBCT) at three positions, lateral (LAT), longitudinal (LNG) and vertical (VRT). IMRT dosimetric parameters of V 5, V 20, and mean dose were collected. Results Patients with BMI ≥28 kg/m2 showed significantly higher absolute values of LAT, LNG and VRT, higher V 5, V 20, mean dose, as well as higher total incidence of radiation pneumonitis and grade III radiation pneumonitis compared with patients with BMI <24 kg/m2 or 24–28 kg/m2. Spearman’s analysis demonstrated that the absolute values of LAT, LNG and VRT were positively correlated with BMI, and positive correlation existed among BMI, dosimetric parameters and setup errors. ROC curves showed that LAT in setup errors and V 5 in dosimetric parameters had the best diagnostic value for prediction of radiation pneumonitis. Only BMI, LAT, V 5 and V 20 were the independent risk factors for radiation pneumonitis. Conclusions Setup error caused by higher BMI might be associated with the dosimetric parameters, as well as the incidence of radiation pneumonitis in NSCLC patients.

Handcrafted Radiomics, Deep Learning Radiomics in the Prediction of Radiation Pneumonitis for NSCLC Patients Treated with Immunotherapy Followed with Thoracic Radiotherapy

Our previous study has shown that NSCLC patients previously received immune checkpoint inhibitors (ICIs) underwent thoracic intensity modulated radiotherapy have a higher risk of acute radiation pneumonitis (RP). This study aimed to establish machine learning models using handcrafted radiomics (HCR), deep learning-based radiomics (DLR) and clinical characteristics to improve the prediction of symptomatic radiation pneumonitis (RP) (grade ≥ 2) status for NSCLC patients treated with immunotherapy followed with thoracic radiotherapy. This study retrospectively collected data of 61 NSCLC patients meeting the requirements of study enrollment. Of these 61 patients, 35 developed symptomatic graded ≥ 2 RP. We defined 3 regions of interest (ROIs) in planning CT images including gross tumor volume (GTV), planning tumor volume (PTV), PTV-GTV. We calculated the mean dose, V5, V10, V20, and V30 within TL-GTV, and the volume of GTV, PTV and total lung. A total of 516 handcrafted radiomics features and 512 deep features were extracted from each 3 ROIs. Person Correlation Analysis and Least Absolute Shrinkage and Selection Operator (LASSO) were used to reduce the dimension of features. The HCR models, DLR models and the fusion models across different ROIs with machine learning classifiers were built and compared. In multi-classifier modeling, models with PTV under logistic regression (LR) classifiers showed better prediction than other ROIs under different machine learning algorithms. Based on PTV with LR, HCR+ DLR model had better performance, with an area under the curve (AUC) of 0.95 (95% confidence interval (CI): 0.893-1) in the training cohort and 0.87 (95% CI: 0.698-1) in the test cohort, which was higher than that of HCR model, with an AUC of 0.86 (95% CI: 0.755-0.9) in the training cohort and 0.82 (95% CI: 0.624-1) in the test cohort, the results of fusion model with HCR, DLR and 7 clinical characteristics including T, N, clinical stage, age, smoking, radiotherapy alone/combined and V30, demonstrated the best distinguishing performance, with an AUC of 0.99 (95% CI: 0.970-1) in the training cohort and 0.91 (95% CI: 0.784-1) in the test cohort. The combination of HCR, DLR and clinical characteristic underwent machine learning algorithms can improve the prediction of symptomatic RP in NSCLC patients treated with ICIs followed with thoracic radiotherapy.

Combining Dosimetric and Radiomics Features for the Prediction of Radiation Pneumonitis in Locally Advanced Non-Small Cell Lung Cancer by Machine Learning

This study aimed to analyze the dosimetric factors and radiomics features of tumor and lungs in locally advanced non-small cell lung cancer (LANSCLC) to establish machine learning models and improve the prediction of grade (G) 2 radiation pneumonitis (RP). This study retrospectively collected data of 284 LANSCLC patients underwent concurrent chemoradiotherapy (CCRT) to a median dose of 64 Gy in 20-33 fractions between 2013 and 2021. Of this cohort, 21.1% of patients had ≥ G2 RP. There were 4 regions of interest (ROIs) had been identified in planning computed tomography images: gross tumor volume (GTV), ipsilesional lung (IL), contralesional lung (CL), and total lung (TL). We calculated the dose-volume histogram (DVH) from the lowest dose to the maximum dose increasing by degrees with 1 Gy, and extracted a total of 172 radiomics features from all the 4 ROIs. We selected the best predictors for classifying 2 groups of patients using a sequential backward elimination support vector machine model. The best predictors for ≥ G2 RP were the combination of 8 radiomics features and 7 dosimetric factors in training group, and the validation group achieved an area under the curve (AUC) of 0.847 (accuracy, 80.38%; sensitivity, 78.95%; specificity, 81.82%). The eight radiomic features included 2 from GTV while 1, 2 and 3 from IL, CL and TL, respectively. For dosimetric factors, V65 of GTV, V20, V50 and V55 of IL, V10 of CL, V20 and V55 of TL appeared to be significantly related to symptomatic RP. These dosimetric factors should be constrained to less than 99.2%, 50.0%, 17.5%, 13.0%, 39.5%, 32.0%, and 6.6%, respectively. Combining dosimetric factors and radiomics features within GTV, IL, CL and TL can improve the prediction of symptomatic RP in LANSCLC patients treated with CCRT. The results suggested the importance of V65 of GTV, V20, V50 and V55 of IL, V10 of CL, V20 and V55 of TL as predictors of symptomatic RP and provide useful information for optimization of treatment planning in the era of combination of radiotherapy and immunotherapy.

Inflammation index predicts radiation-induced lung injury and prognosis in lung tumors treated with stereotactic body radiation therapy.

To investigate the effect of inflammation-based indexes in predicting radiation pneumonitis (RP) and prognosis in lung tumor patients treated with stereotactic body radiation therapy (SBRT). The data of one hundred and seventy-two patients with 272 lung lesions from November 2015 to December 2020 were retrospectively analyzed. Pretreatment hematological indexes including platelet count, neutrophil count, and lymphocyte count were collected before treatment. Systemic immune-inflammation index (SII), neutrophil-to-lymphocyte ratio (NLR), and platelet-to-lymphocyte ratio (PLR) were calculated. The receiver operating characteristic (ROC) curve was established to predict the RP and overall survival of patients, and the Youden index was calculated to determine the cutoff values of SII, NLR, and PLR before radiotherapy. Pretreatment SII, NLR, and PLR could predict RP in lung tumor patients treated with SBRT, the optimal cutoff values of SII, NLR, and PLR were 355.38, 2.04, and 141.09, respectively. Pretreatment PLR could predict survival and the optimal cutoff value of PLR was 166.83, patients with a PLR > 166.83 predict worse overall survival (OS) (P < 0.001). The 1-year and 2-year OS for patients with a PLR ≤ 166.83 were 96.3% and 82.4%, while for those with a PLR > 166.83 were 82.0% and 58.5%, respectively. In lung tumor patients treated with SBRT, pretreatment SII, NLR, and PLR can effectively predict RP and a higher PLR predicts poor OS. These inflammation-based indexes could serve as reliable and convenient predictors to guide treatment for physicians in clinical practice.

Automation Radiomics in Predicting Radiation Pneumonitis (RP)

Radiomics has shown great promise in predicting various diseases. Researchers have previously attempted to include radiomics in their automated detection, diagnosis, and segmentation algorithms, taking these steps based on the promising outcomes of radiomics-based studies. As a result of the increased attention given to this topic, numerous institutions have developed their own radiomics software. These packages, on the other hand, have been utilized interchangeably without regard for their fundamental differences. The primary purpose of this study was to explore benefits of predictive model performance for radiation pneumonitis (RP), which is the most frequent side effect of chest radiotherapy, and through this work, we developed a radiomics model based on deep learning that intends to increase RP prediction performance by combining more data points and digging deeper into these data. In order to evaluate the most popular machine learning models, radiographic characteristics were used, and we recorded the most important of them. The high dimensionality of radiomic datasets is a major issue. The method proposed for use in data problems is the synthetic minority oversampling technique, which we used in order to create a balanced dataset by leveraging suitable hardware and open-source software. The present study assessed the efficacy of various machine learning models, including logistic regression (LR), support vector machine (SVM), random forest (RF), and deep neural network (DNN), in predicting radiation pneumonitis by utilizing specific radiomics features. The findings of the study indicate that the four models displayed satisfactory efficacy in forecasting radiation pneumonitis. The DNN model demonstrated the highest area under the receiver operating curve (AUC-ROC) value, which was 0.87, suggesting its superior predictive capacity among the models considered. The AUC-ROC values for the random forest, SVM, and logistic regression models were 0.85, 0.83, and 0.81, respectively.

PO-2115 federated learning of decentralized AI for predicting radiation pneumonitis in lung cancer patients

PO-2115 federated learning of decentralized AI for predicting radiation pneumonitis in lung cancer patients

Computed tomography and radiation dose images-based deep-learning model for predicting radiation pneumonitis in lung cancer patients after radiation therapy

PurposeTo develop a deep learning model that combines CT and radiation dose (RD) images to predict the occurrence of radiation pneumonitis (RP) in lung cancer patients who received radical (chemo)radiotherapy. MethodsCT, RD images and clinical parameters were obtained from 314 retrospectively-collected patients (training set) and 35 prospectively-collected patients (test-set-1) who were diagnosed with lung cancer and received radical radiotherapy in the dose range of 50 Gy and 70 Gy. Another 194 (60 Gy group, test-set-2) and 158 (74 Gy group, test-set-3) patients from the clinical trial RTOG 0617 were used for external validation. A ResNet architecture was used to develop a prediction model that combines CT and RD features. Thereafter, the CT and RD weights were adjusted by using 40 patients from test-set-2 or 3 to accommodate cohorts with different clinical settings or dose delivery patterns. Visual interpretation was implemented using a gradient-weighted class activation map (grad-CAM) to observe the area of model attention during the prediction process. To improve the usability, ready-to-use online software was developed. ResultsThe discriminative ability of a baseline trained model had an AUC of 0.83 for test-set-1, 0.55 for test-set-2, and 0.63 for test-set-3. After adjusting CT and RD weights of the model using a subset of the RTOG-0617 subjects, the discriminatory power of test-set-2 and 3 improved to AUC 0.65 and AUC 0.70, respectively. Grad-CAM showed the regions of interest to the model that contribute to the prediction of RP. ConclusionA novel deep learning approach combining CT and RD images can effectively and accurately predict the occurrence of RP, and this model can be adjusted easily to fit new cohorts.

Ipsilateral lung dose as a correlative measure for radiation pneumonitis in patients treated with definitive concurrent radiochemotherapy.

Mean lung dose (MLD) and percent of total lung (TL) volume that receive a dose greater than 20 Gy (V20) have been the most validated parameters in the prediction of radiation pneumonitis (RP). However, these parameters present mean values of TL parenchyma and predict the right and the left lung as a unique functional organ unit, not take into account the difference in function and dose density between the lungs. Furthermore, there have been very limited data evaluating ipsilateral lung dosimetric constraints in addition to TL parameters to predict RP in non-small cell lung cancer (NSCLC) patients treated with radiochemotherapy (RCT). Between 2010 and 2017, clinical-radiological findings of NSCLC patients treated with RCT were evaluated in terms of RP, retrospectively. MLD, V20, and V30 values of ipsilateral lung were assessed from dose-volume histogram and registered. The primary endpoint was to assess the relation between ipsilateral lung dose constraints and RP risk. There were 75 patients. There was ≥Grade 2 RP in 33 cases (%44). In univariate analysis, ipsilateral MLD, ipsilateral V20, ipsilateral V30, and TL V30 were found to be significant. Ipsilateral MLD and PTV were found to be the independent risk factors for RP. Cutoff values for RP risk were determined as 18Gy, 35%, and 28% for ipsilateral MLD, ipsilateral V20, and ipsilateral V30, respectively. Predictive values for ipsilateral MLD and ipsilateral V20 were higher than TL. In NSCLC patients treated with RCT, MLD, V20, and V30 values of ipsilateral lung parameters might increase the predictability of RP risk in addition to TL parameters. Cutoff values for RP risk were determined as 18Gy, 35%, and 28% for ipsilateral MLD, ipsilateral V20, and ipsilateral V30, respectively. Predictive values for ipsilateral MLD and ipsilateral V20 were higher than TL.

3D deep convolution neural network for radiation pneumonitis prediction following stereotactic body radiotherapy.

In this study, we investigated 3D convolutional neural networks (CNNs) with input from radiographic and dosimetric datasets of primary lung tumors and surrounding lung volumes to predict the likelihood of radiation pneumonitis (RP). Pre-treatment, 3- and 6-month follow-up computed tomography (CT) and 3D dose datasets from one hundred and ninety-three NSCLC patients treated with stereotactic body radiotherapy (SBRT) were retrospectively collected and analyzed for this study. DenseNet-121 and ResNet-50 models were selected for this study as they are deep neural networks and have been proven to have high accuracy for complex image classification tasks. Both were modified with 3D convolution and max pooling layers to accept 3D datasets. We used a minority class oversampling approach and data augmentation to address the challenges of data imbalance and data scarcity. We built two sets of models for classification of three (No RP, Grade 1 RP, Grade 2 RP) and two (No RP, Yes RP) classes as outputs. The 3D DenseNet-121 models performed better (F1 score [0.81], AUC [0.91] [three class]; F1 score [0.77], AUC [0.84] [two class]) than the 3D ResNet-50 models (F1 score [0.54], AUC [0.72] [three-class]; F1 score [0.68], AUC [0.71] [two-class]) (p=0.017 for three class predictions). We also attempted to identify salient regions within the input 3D image dataset via integrated gradient (IG) techniques to assess the relevance of the tumor surrounding volume for RP stratification. These techniques appeared to indicate the significance of the tumor and surrounding regions in the prediction of RP. Overall, 3D CNNs performed well to predict clinical RP in our cohort based on the provided image sets and radiotherapy dose information.

Radiation pneumonitis prediction model with integrating multiple dose-function features on 4DCT ventilation images

PurposeRadiation pneumonitis (RP) is dose-limiting toxicity for non-small-cell cancer (NSCLC). This study developed an RP prediction model by integrating dose-function features from computed four-dimensional computed tomography (4DCT) ventilation using the least absolute shrinkage and selection operator (LASSO). MethodsBetween 2013 and 2020, 126 NSCLC patients were included in this study who underwent a 4DCT scan to calculate ventilation images. We computed two sets of candidate dose-function features from (1) the percentage volume receiving > 20 Gy or the mean dose on the functioning zones determined with the lower cutoff percentile ventilation value, (2) the functioning zones determined with lower and upper cutoff percentile ventilation value using 4DCT ventilation images. An RP prediction model was developed by LASSO while simultaneously determining the regression coefficient and feature selection through fivefold cross-validation. ResultsWe found 39.3 % of our patients had a ≥ grade 2 RP. The mean area under the curve (AUC) values for the developed models using clinical, dose-volume, and dose-function features with a lower cutoff were 0.791, and the mean AUC values with lower and upper cutoffs were 0.814. The relative regression coefficient (RRC) on dose-function features with upper and lower cutoffs revealed a relative impact of dose to each functioning zone to RP. RRCs were 0.52 for the mean dose on the functioning zone, with top 20 % of all functioning zone was two times greater than that of 0.19 for these with 60 %–80 % and 0.17 with 40 %–60 % (P < 0.01). ConclusionsThe introduction of dose-function features computed from functioning zones with lower and upper cutoffs in a machine learning framework can improve RP prediction. The RRC given by LASSO using dose-function features allows for the quantification of the RP impact of dose on each functioning zones and having the potential to support treatment planning on functional image-guided radiotherapy.