Application of radiomics signature captured from pretreatment thoracic CT to predict brain metastases in stage III/IV ALK-positive non-small cell lung cancer patients

Potential of radiomics analysis and machine learning for predicting brain metastasis in newly diagnosed lung cancer patients

AIM

To explore the potential of utilising radiomics analysis and machine-learning models that incorporate intratumoural and peritumoural regions of interest (ROIs) for predicting brain metastasis (BM) in newly diagnosed lung cancer patients.

MATERIALS AND METHODS

The study comprised 183 lung cancer patients (training cohort: n=146; validation cohort: n=37) whose radiomics features were extracted from plain computed tomography (CT) images of the primary lesion. Four machine-learning algorithms (logistic regression [LR], support vector machine [SVM], k-nearest neighbour algorithm [KNN], and random forest [RF]) were employed to develop predictive models. Model diagnostic performance was assessed through receiver operating characteristic (ROC) analysis, and clinical utility was evaluated using decision curve analysis (DCA). Finally, the radiomics model's generalisation ability was further validated in the prediction of metachronous brain metastasis (MBM).

RESULTS

After feature screening, 22 radiomics features were identified as highly predictive, of which nine were derived from the peritumour region. All four machine-learning models demonstrated predictive capability, with SVM showing superior efficiency and robustness. The area under the ROC curve (AUC) of SVM was 0.918 in the training cohort and 0.901 in the validation cohort. DCA indicated the highest net benefit. Furthermore, the time-dependent ROC curve exhibited predictive efficacy for MBM occurrence across 1-, 2-, and 3-year follow-up periods, with all AUC values exceeding 0.7.

CONCLUSION

The optimal SVM model integrating intratumoural and peritumoural radiomics features was confirmed and defined as an imaging biomarker for predicting BM in newly diagnosed lung cancer patients, underscoring its potential to significantly impact clinical diagnosis and treatment.

Enhancing brain metastasis prediction in non-small cell lung cancer: a deep learning-based segmentation and CT radiomics-based ensemble learning model

BACKGROUND

Brain metastasis (BM) is most common in non-small cell lung cancer (NSCLC) patients. This study aims to enhance BM risk prediction within three years for advanced NSCLC patients by using a deep learning-based segmentation and computed tomography (CT) radiomics-based ensemble learning model.

METHODS

This retrospective study included 602 stage IIIA-IVB NSCLC patients, 309 BM patients and 293 non-BM patients, from two centers. Patients were divided into a training cohort (N = 376), an internal validation cohort (N = 161) and an external validation cohort (N = 65). Lung tumors were first segmented by using a three-dimensional (3D) deep residual U-Net network. Then, a total of 1106 radiomics features were computed by using pretreatment lung CT images to decode the imaging phenotypes of primary lung cancer. To reduce the dimensionality of the radiomics features, recursive feature elimination configured with the least absolute shrinkage and selection operator (LASSO) regularization method was applied to select the optimal image features after removing the low-variance features. An ensemble learning algorithm of the extreme gradient boosting (XGBoost) classifier was used to train and build a prediction model by fusing radiomics features and clinical features. Finally, Kaplan‒Meier (KM) survival analysis was used to evaluate the prognostic value of the prediction score generated by the radiomics-clinical model.

RESULTS

The fused model achieved area under the receiver operating characteristic curve values of 0.91 ± 0.01, 0.89 ± 0.02 and 0.85 ± 0.05 on the training and two validation cohorts, respectively. Through KM survival analysis, the risk score generated by our model achieved a significant prognostic value for BM-free survival (BMFS) and overall survival (OS) in the two cohorts (P < 0.05).

CONCLUSIONS

Our results demonstrated that (1) the fusion of radiomics and clinical features can improve the prediction performance in predicting BM risk, (2) the radiomics model generates higher performance than the clinical model, and (3) the radiomics-clinical fusion model has prognostic value in predicting the BMFS and OS of NSCLC patients.

Radiomics and artificial intelligence for precision medicine in lung cancer treatment

Lung cancer is the leading cause of cancer-related deaths worldwide. It exhibits, at the mesoscopic scale, phenotypic characteristics that are generally indiscernible to the human eye but can be captured non-invasively on medical imaging as radiomic features, which can form a high dimensional data space amenable to machine learning. Radiomic features can be harnessed and used in an artificial intelligence paradigm to risk stratify patients, and predict for histological and molecular findings, and clinical outcome measures, thereby facilitating precision medicine for improving patient care. Compared to tissue sampling-driven approaches, radiomics-based methods are superior for being non-invasive, reproducible, cheaper, and less susceptible to intra-tumoral heterogeneity. This review focuses on the application of radiomics, combined with artificial intelligence, for delivering precision medicine in lung cancer treatment, with discussion centered on pioneering and groundbreaking works, and future research directions in the area.

The Association of Gross Tumor Volume and Its Radiomics Features with Brain Metastases Development in Patients with Radically Treated Stage III Non-Small Cell Lung Cancer

PURPOSE

To identify clinical risk factors, including gross tumor volume (GTV) and radiomics features, for developing brain metastases (BM) in patients with radically treated stage III non-small cell lung cancer (NSCLC).

METHODS

Clinical data and planning CT scans for thoracic radiotherapy were retrieved from patients with radically treated stage III NSCLC. Radiomics features were extracted from the GTV, primary lung tumor (GTVp), and involved lymph nodes (GTVn), separately. Competing risk analysis was used to develop models (clinical, radiomics, and combined model). LASSO regression was performed to select radiomics features and train models. Area under the receiver operating characteristic curves (AUC-ROC) and calibration were performed to assess the models' performance.

RESULTS

Three-hundred-ten patients were eligible and 52 (16.8%) developed BM. Three clinical variables (age, NSCLC subtype, and GTVn) and five radiomics features from each radiomics model were significantly associated with BM. Radiomic features measuring tumor heterogeneity were the most relevant. The AUCs and calibration curves of the models showed that the GTVn radiomics model had the best performance (AUC: 0.74; 95% CI: 0.71-0.86; sensitivity: 84%; specificity: 61%; positive predictive value [PPV]: 29%; negative predictive value [NPV]: 95%; accuracy: 65%).

CONCLUSION

Age, NSCLC subtype, and GTVn were significant risk factors for BM. GTVn radiomics features provided higher predictive value than GTVp and GTV for BM development. GTVp and GTVn should be separated in clinical and research practice.

Computed-tomography-based radiomic nomogram for predicting the risk of indeterminate small (5-20 mm) solid pulmonary nodules

PURPOSE

This study aims to develop a diagnostic model that combines computed tomography (CT) images and radiomic features to differentiate indeterminate small (5-20 mm) solid pulmonary nodules (SSPNs).

METHODS

This study retrospectively enrolled 413 patients who had had SSPNs surgically removed and histologically confirmed between 2017 and 2019. The SSPNs included solid malignant pulmonary nodules (n = 210) and benign pulmonary nodules (n = 203). The least absolute shrinkage and selection operator was used for radiomic feature selection, and random forest algorithms were used for radiomic model construction. The clinical model and nomogram were established using univariate and multivariable logistic regression analyses combined with clinical symptoms, subjective CT findings, and radiomic features. The area under the curve (AUC) of the receiver operating characteristic curve was used to evaluate the performance of the models.

RESULTS

The AUC for the clinical model was 0.77 in the training cohort [n = 289; 95% confidence interval (CI): 0.71-0.82; P = 0.001] and 0.75 in the validation cohort (n = 124; 95% CI: 0.66-0.83; P = 0.016). The AUCs for the nomogram were 0.92 (95% CI: 0.89-0.95; P < 0.001) and 0.85 (95% CI: 0.78-0.91; P < 0.001), respectively. The radiomic score (Rad-score), sex, pleural indentation, and age were the independent predictors that were used to build the nomogram.

CONCLUSION

The radiomic nomogram derived from clinical features, subjective CT signs, and the Rad-score can potentially identify the risk of indeterminate SSPNs and aid in the patient's preoperative diagnosis.

A "Seed-and-Soil" Radiomics Model Predicts Brain Metastasis Development in Lung Cancer: Implications for Risk-Stratified Prophylactic Cranial Irradiation

Introduction: Brain is a major site of metastasis for lung cancer, and effective therapy for developed brain metastasis (BM) is limited. Prophylactic cranial irradiation (PCI) has been shown to reduce BM rate and improve survival in small cell lung cancer, but this result was not replicated in unselected non-small cell lung cancer (NSCLC) and had the risk of inducing neurocognitive dysfunctions. We aimed to develop a radiomics BM prediction model for BM risk stratification in NSCLC patients. Methods: 256 NSCLC patients with no BM at baseline brain magnetic resonance imaging (MRI) were selected; 128 patients developed BM within three years after diagnosis and 128 remained BM-free. For radiomics analysis, both the BM and non-BM groups were randomly distributed into training and testing datasets at an 70%:30% ratio. Both brain MRI (representing the soil) and chest computed tomography (CT, representing the seed) radiomic features were extracted to develop the BM prediction models. We first developed the radiomic models using the training dataset (89 non-BM and 90 BM cases) and subsequently validated the models in the testing dataset (39 non-BM and 38 BM cases). A radiomics BM score (RadBM score) was generated, and BM-free survival were compared between RadBM score-high and RadBM score-low groups. Results: The radiomics model developed from baseline brain MRI features alone can predict BM development in NSCLC patients. A fusion model integrating brain MRI features with primary tumor CT features (seed-and-soil model) provided synergetic effect and was more efficient in predicting BM (areas under the receiver operating characteristic curve 0.84 (95% confidence interval: 0.80−0.89) and 0.80 (95% confidence interval: 0.71−0.88) in the training and testing datasets, respectively). BM-free survival was significantly shorter in the RadBM score-high group versus the RadBM score-low group (Log-rank, p < 0.001). Hazard ratios for BM were 1.056 (95% confidence interval: 1.044−1.068) per 0.01 increment in RadBM score. Cumulative BM rates at three years were 75.8% and 24.2% for the RadBM score-high and RadBM score-low groups, respectively. Only 1.2% (7/565) of the BM lesions were located within the hippocampal avoidance region. Conclusion: The results demonstrated that intrinsic features of a non-metastatic brain exert a significant impact on BM development, which is first-in-class in metastasis prediction studies. A radiomics BM prediction model utilizing both primary tumor and pre-metastatic brain features might provide a useful tool for individualized PCI administration in NSCLC patients more prone to develop BM.

Feature selection methods and predictive models in CT lung cancer radiomics

Radiomics is a technique that extracts quantitative features from medical images using data-characterization algorithms. Radiomic features can be used to identify tissue characteristics and radiologic phenotyping that is not observable by clinicians. A typical workflow for a radiomics study includes cohort selection, radiomic feature extraction, feature and predictive model selection, and model training and validation. While there has been increasing attention given to radiomic feature extraction, standardization, and reproducibility, currently, there is a lack of rigorous evaluation of feature selection methods and predictive models. Herein, we review the published radiomics investigations in CT lung cancer and provide an overview of the commonly used radiomic feature selection methods and predictive models. We also compare limitations of various methods in clinical applications and present sources of uncertainty associated with those methods. This review is expected to help raise awareness of the impact of radiomic feature and model selection methods on the integrity of radiomics studies.

Investigation of the added value of CT-based radiomics in predicting the development of brain metastases in patients with radically treated stage III NSCLC

Introduction:

Despite radical intent therapy for patients with stage III non-small-cell lung cancer (NSCLC), cumulative incidence of brain metastases (BM) reaches 30%. Current risk stratification methods fail to accurately identify these patients. As radiomics features have been shown to have predictive value, this study aims to develop a model combining clinical risk factors with radiomics features for BM development in patients with radically treated stage III NSCLC.

Methods:

Retrospective analysis of two prospective multicentre studies. Inclusion criteria: adequately staged [¹⁸F-fluorodeoxyglucose positron emission tomography-computed tomography (18-FDG-PET-CT), contrast-enhanced chest CT, contrast-enhanced brain magnetic resonance imaging/CT] and radically treated stage III NSCLC, exclusion criteria: second primary within 2 years of NSCLC diagnosis and prior prophylactic cranial irradiation. Primary endpoint was BM development any time during follow-up (FU). CT-based radiomics features (N = 530) were extracted from the primary lung tumour on 18-FDG-PET-CT images, and a list of clinical features (N = 8) was collected. Univariate feature selection based on the area under the curve (AUC) of the receiver operating characteristic was performed to identify relevant features. Generalized linear models were trained using the selected features, and multivariate predictive performance was assessed through the AUC.

Results:

In total, 219 patients were eligible for analysis. Median FU was 59.4 months for the training cohort and 67.3 months for the validation cohort; 21 (15%) and 17 (22%) patients developed BM in the training and validation cohort, respectively. Two relevant clinical features (age and adenocarcinoma histology) and four relevant radiomics features were identified as predictive. The clinical model yielded the highest AUC value of 0.71 (95% CI: 0.58–0.84), better than radiomics or a combination of clinical parameters and radiomics (both an AUC of 0.62, 95% CIs of 0.47–076 and 0.48–0.76, respectively).

Conclusion:

CT-based radiomics features of primary NSCLC in the current setup could not improve on a model based on clinical predictors (age and adenocarcinoma histology) of BM development in radically treated stage III NSCLC patients.

Pretreatment Thoracic CT Radiomic Features to Predict Brain Metastases in Patients With ALK-Rearranged Non-Small Cell Lung Cancer

Objective: To identify CT imaging biomarkers based on radiomic features for predicting brain metastases (BM) in patients with ALK-rearranged non-small cell lung cancer (NSCLC).

Methods: NSCLC patients with pathologically confirmed ALK rearrangement from January 2014 to December 2020 in our hospital were enrolled retrospectively in this study. Finally, 77 patients were included according to the inclusion and exclusion criteria. Patients were divided into two groups: BM+ were those patients who were diagnosed with BM at baseline examination (n = 16) or within 1 year’s follow-up (n = 14), and BM− were those without BM followed up for at least 1 year (n = 47). Radiomic features were extracted from the pretreatment thoracic CT images. Sequential univariate logistic regression, LASSO regression, and backward stepwise logistic regression were used to select radiomic features and develop a BM-predicting model.

Results: Five robust radiomic features were found to be independent predictors of BM. AUC for radiomics model was 0.828 (95% CI: 0.736–0.921), and when combined with clinical features, the AUC was increased (p = 0.017) to 0.909 (95% CI: 0.845–0.972). The individualized BM-predicting model incorporated with clinical features was visualized by the nomogram.

Conclusion: Radiomic features extracted from pretreatment thoracic CT images have the potential to predict BM within 1 year after detection of the primary tumor in patients with ALK-rearranged NSCLC. The radiomics model incorporated with clinical features shows improved risk stratification for such patients.

Development and validation a radiomics nomogram for diagnosing occult brain metastases in patients with stage IV lung adenocarcinoma

Background

To develop and validate a radiomics model using computed tomography (CT) images acquired from the first diagnosis to estimate the status of occult brain metastases (BM) in patients with stage IV lung adenocarcinoma (LADC).

Methods

One hundred and ninety-three patients who were first diagnosed with stage IV LADC were enrolled and divided into a training cohort (n=135) and a validation cohort (n=58). Then, 725 radiomic features were extracted from contoured primary tumor volumes of LADCs. Intra- and interobserver reliabilities were calculated, and the least absolute shrinkage and selection operator (LASSO) was applied for feature selection. Subsequently, a radiomics signature (Rad-Score) was built. To improve performance, a nomogram incorporating a radiomics signature and an independent clinical predictor was developed. Finally, the established signature and nomogram were assessed using receiver operating characteristic (ROC) curves and precision-recall curves (PRC). Both empirical and α-binomial model-based ROCs and PRCs were plotted, and the area under the curve (AUC) and average precision (AP) of ROCs and PRCs were calculated and compared.

Results

A radiomics signature and Rad-Score were constructed using eight radiomic features, and these had significant correlations with occult BM status. A nomogram was developed by incorporating a Rad-Score and the primary tumor location. The nomogram yielded an optimal AUC of 0.911 [95% confidence interval (CI): 0.903–0.919] and an AP of 0.885 (95% CI: 0.876–0.894) in the training cohort, and an AUC of 0.873 (95% CI: 0.866–0.80) and an AP of 0.827 (95% CI: 0.820–0.834) in the validation cohort using α-binomial model-based method. The calibration curve demonstrated that the nomogram showed high agreement between the actual occult BM probability and predicted by the nomogram (P=0.427).

Conclusions

The nomogram incorporating a radiomics signature and a clinical risk factor achieved optimal performance after holistic assessment using unbiased indexes for diagnosing occult BM of patients who were first diagnosed with stage IV LADC.

Incorporating radiomics into clinical trials: expert consensus endorsed by the European Society of Radiology on considerations for data-driven compared to biologically driven quantitative biomarkers

Existing quantitative imaging biomarkers (QIBs) are associated with known biological tissue characteristics and follow a well-understood path of technical, biological and clinical validation before incorporation into clinical trials. In radiomics, novel data-driven processes extract numerous visually imperceptible statistical features from the imaging data with no a priori assumptions on their correlation with biological processes. The selection of relevant features (radiomic signature) and incorporation into clinical trials therefore requires additional considerations to ensure meaningful imaging endpoints. Also, the number of radiomic features tested means that power calculations would result in sample sizes impossible to achieve within clinical trials. This article examines how the process of standardising and validating data-driven imaging biomarkers differs from those based on biological associations. Radiomic signatures are best developed initially on datasets that represent diversity of acquisition protocols as well as diversity of disease and of normal findings, rather than within clinical trials with standardised and optimised protocols as this would risk the selection of radiomic features being linked to the imaging process rather than the pathology. Normalisation through discretisation and feature harmonisation are essential pre-processing steps. Biological correlation may be performed after the technical and clinical validity of a radiomic signature is established, but is not mandatory. Feature selection may be part of discovery within a radiomics-specific trial or represent exploratory endpoints within an established trial; a previously validated radiomic signature may even be used as a primary/secondary endpoint, particularly if associations are demonstrated with specific biological processes and pathways being targeted within clinical trials. KEY POINTS: • Data-driven processes like radiomics risk false discoveries due to high-dimensionality of the dataset compared to sample size, making adequate diversity of the data, cross-validation and external validation essential to mitigate the risks of spurious associations and overfitting. • Use of radiomic signatures within clinical trials requires multistep standardisation of image acquisition, image analysis and data mining processes. • Biological correlation may be established after clinical validation but is not mandatory.

The development and validation of a nomogram for predicting brain metastases in lung squamous cell carcinoma patients: an analysis of the Surveillance, Epidemiology, and End Results (SEER) database

Background

The incidence of brain metastasis (BM) in patients suffering from lung squamous cell carcinoma (LUSC) is lower than that in patients suffering from non-squamous cell carcinoma (NSCC) and there are few studies on BM of LUSC. The purpose of this investigation was to ascertain the risk factors of LUSC, as well as to establish a nomogram prognostic model to predict the incidence of BM in patients with LUSC.

Methods

Patients diagnosed with LUSC between 2010 and 2015 were identified from the Surveillance, Epidemiology, and End Results (SEER) database and the patient data were collated. All patients diagnosed from 2010–2012 were allocated into the training cohort, and the remaining patients diagnosed from 2013–2015 formed the test cohort. Using factors that were screened out through logistic regression analyses, the nomogram in the training cohort was established. It was then evaluated for discrimination and calibration using the test cohort. The performance of the nomogram was assessed by quantifying the area under the receiver operating characteristic (ROC) curve and evaluating the calibration curve.

Results

A total of 26,154 LUSC patients were included in the study. The training cohort consisted of 16,543 patients and there were 8611 patients in the test cohort. Age, marital status, insurance status, histological grade, tumor location, laterality, stage of the cancer, number of metastatic organs, chemotherapy, surgery, and radiotherapy were highly correlated with the incidence of BM. The area under the ROC curve (AUC) of the nomogram for the training cohort and the test cohort were 0.810 [95% confidence interval (CI): 0.796 to 0.823] and 0.805 (95% CI: 0.784 to 0.825), respectively. The slope of the calibration curve was close to 1.

Conclusions

The nomogram was able to accurately predict the incidence of BM. This may be beneficial for the early identification of high-risk LUSC patients and the establishment of individualized treatments.

Radiomics: an overview in lung cancer management-a narrative review

Radiomics is a novel approach for optimizing the analysis massive data from medical images to provide auxiliary guidance in clinical issues. Quantitative feature extraction is one of the critical steps of radiomics. The association between radiomics features and the clinicopathological information of diseases can be identified by several statistics methods. For instance, although significant progress has been made in the field of lung cancer, too many questions remain, especially for the individualized decisions. Radiomics offers a new tool to encode the characteristics of lung cancer which is the leading cause of cancer-related deaths worldwide. Here, we reviewed the workflow and clinical utility of radiomics in lung cancer management, including pulmonary nodules detection, classification, histopathology and genetics evaluation, clinical staging, therapy response, and prognosis prediction. Most of these studies showed positive results, indicating the potential value of radiomics in clinical practice. The implementation of radiomics is both feasible and invaluable, and has aided clinicians in ascertaining the nature of a disease with greater precision. However, it should be noted that radiomics in its current state cannot completely replace the work of therapists or tissue examination. The potential future trends of this modality were also remarked. More efforts are needed to overcome the limitations identified above in order to facilitate the widespread application of radiomics in the reasonably near future.

Radiomics as a personalized medicine tool in lung cancer: Separating the hope from the hype

Radiomics has become a popular image analysis method in the last few years. Its key hypothesis is that medical images harbor biological, prognostic and predictive information that is not revealed upon visual inspection. In contrast to previous work with a priori defined imaging biomarkers, radiomics instead calculates image features at scale and uses statistical methods to identify those most strongly associated to outcome. This builds on years of research into computer aided diagnosis and pattern recognition. While the potential of radiomics to aid personalized medicine is widely recognized, several technical limitations exist which hinder biomarker translation. Aspects of the radiomic workflow lack repeatability or reproducibility under particular circumstances, which is a key requirement for the translation of imaging biomarkers into clinical practice. One of the most commonly studied uses of radiomics is for personalized medicine applications in Non-Small Cell Lung Cancer (NSCLC). In this review, we summarize reported methodological limitations in CT based radiomic analyses together with suggested solutions. We then evaluate the current NSCLC radiomics literature to assess the risk associated with accepting the published conclusions with respect to these limitations. We review different complementary scoring systems and initiatives that can be used to critically appraise data from radiomics studies. Wider awareness should improve the quality of ongoing and future radiomics studies and advance their potential as clinically relevant biomarkers for personalized medicine in patients with NSCLC.

Multiclassifier fusion based on radiomics features for the prediction of benign and malignant primary pulmonary solid nodules

Background

To test the ability of a multiclassifier model based on radiomics features to predict benign and malignant primary pulmonary solid nodules.

Methods

Computed tomography (CT) images of 342 patients with primary pulmonary solid nodules confirmed by histopathology or follow-up were retrospectively analyzed. The region of interest (ROI) of the images was delineated, and the radiomics features of the lesions were extracted. The feature weight was calculated using the relief feature selection algorithm. Based on the selected features, five classifier models were constructed: support vector machine (SVM), random forest (RF), logistic regression (LR), extreme learning machine (ELM), and K-nearest neighbor (KNN). The precision, recall rate, and area under the receiver operating characteristic curve (AUC) were used to evaluate the prediction performance of each classifier. The prediction result of each classifier was first weighted, and then all the prediction results were fused to predict the nodule type of unknown images. The prediction precision, recall rate, and AUC of the fusion classifier and single classifier were compared. Cross-validation was used to evaluate the generalization of the fusion classifier, and t- and F-tests were performed on the five classifiers and fusion classifier.

Results

For each ROI, 450 features in four major categories were extracted and were analyzed using the relief feature selection algorithm. According to the weights, 25 highly repetitive and nonredundant stable features that played a major role in pulmonary nodule classification were selected. The fusion classifier’s prediction performance (prediction precision =92.0%, AUC =0.915) was superior to those of SVM (prediction precision =75.3%, AUC =0.740), RF (prediction precision =89.1%, AUC =0.855), LR (prediction precision =68.4%, AUC =0.681), ELM (prediction precision =87.0%, AUC =0.830), and KNN (prediction precision =77.1%, AUC =0.702). The fusion classifier showed the best null hypothesis performance in the t-test (P=0.035) and F-test (P=0.036).

Conclusions

The multiclassifier fusion model based on radiomics features had high prediction value for benign and malignant primary pulmonary solid nodules.