CT Slice Thickness and Convolution Kernel Affect Performance of a Radiomic Model for Predicting EGFR Status in Non-Small Cell Lung Cancer: A Preliminary Study

Integrating radiomics and deep learning for enhanced prediction of high-grade patterns in stage IA lung adenocarcinoma

Background

The presence of high-grade patterns (HGPs) often has a detrimental effect on prognosis. It is helpful to make individualized clinical treatment plans when preoperative recognition of the presence of HGPs becomes possible. So, this study aimed to develop a model based on preoperative computed tomography (CT) images to predict the presence of HPGs in invasive pulmonary non-mucinous adenocarcinoma.

Methods

A retrospective analysis was conducted on 403 surgically treated patients with clinical stage IA and pathologically confirmed invasive non-mucinous adenocarcinoma. There were 256 deep learning features and 1,836 handcrafted features extracted from the regions of interest (ROIs) in preoperative CT images. The optimal subset of features was screened using t-test, Pearson correlation analysis, and least absolute shrinkage and selection operator (LASSO) regression to construct the fusion model. Receiver operating characteristic (ROC) curve was applied to assess the model's performance. Decision curve analysis (DCA) and calibration curve were used to assess the clinical usefulness.

Results

The fusion model combining radiomics features and deep learning features using the XGBoost classifier exhibited strong predictive efficacy with the area under the curve (AUC) of 0.983, 0.862, and 0.832 in the training, validation, and test set. It means that the model can distinguish well between tumors with and without HGPs. The fusion model had better diagnostic performance when compared to the radiomics model and deep learning model. Calibration curve indicated good coherence between model prediction and the actual observation. DCA revealed the fusion model exerted the highest clinical benefit.

Conclusions

The fusion model can identify the presence of HPGs in invasive lung adenocarcinoma from preoperative CT images. It assists clinicians in determining individualized treatments and monitoring strategies for patients.

Performance of Lung Cancer Prediction Models for Screening-detected, Incidental, and Biopsied Pulmonary Nodules

Purpose To evaluate the performance of eight lung cancer prediction models on patient cohorts with screening-detected, incidentally detected, and bronchoscopically biopsied pulmonary nodules. Materials and Methods This study retrospectively evaluated promising predictive models for lung cancer prediction in three clinical settings: lung cancer screening with low-dose CT, incidentally detected pulmonary nodules, and nodules deemed suspicious enough to warrant a biopsy. The area under the receiver operating characteristic curve of eight validated models, including logistic regressions on clinical variables and radiologist nodule characterizations, artificial intelligence (AI) on chest CT scans, longitudinal imaging AI, and multimodal approaches for prediction of lung cancer risk was assessed in nine cohorts (n = 898, 896, 882, 219, 364, 117, 131, 115, 373) from multiple institutions. Each model was implemented from their published literature, and each cohort was curated from primary data sources collected over periods from 2002 to 2021. Results No single predictive model emerged as the highest-performing model across all cohorts, but certain models performed better in specific clinical contexts. Single-time-point chest CT AI performed well for screening-detected nodules but did not generalize well to other clinical settings. Longitudinal imaging and multimodal models demonstrated comparatively good performance on incidentally detected nodules. When applied to biopsied nodules, all models showed low performance. Conclusion Eight lung cancer prediction models failed to generalize well across clinical settings and sites outside of their training distributions. Keywords: Diagnosis, Classification, Application Domain, Lung Supplemental material is available for this article. © RSNA, 2025 See also commentary by Shao and Niu in this issue.

A comparative analysis of image harmonization techniques in mitigating differences in CT acquisition and reconstruction

Objective. The study aims to systematically characterize the effect of CT parameter variations on images and lung radiomic and deep features, and to evaluate the ability of different image harmonization methods to mitigate the observed variations.Approach. A retrospective in-house sinogram dataset of 100 low-dose chest CT scans was reconstructed by varying radiation dose (100%, 25%, 10%) and reconstruction kernels (smooth, medium, sharp). A set of image processing, convolutional neural network (CNNs), and generative adversarial network-based (GANs) methods were trained to harmonize all image conditions to a reference condition (100% dose, medium kernel). Harmonized scans were evaluated for image similarity using peak signal-to-noise ratio (PSNR), structural similarity index measure (SSIM), and learned perceptual image patch similarity (LPIPS), and for the reproducibility of radiomic and deep features using concordance correlation coefficient (CCC).Main Results. CNNs consistently yielded higher image similarity metrics amongst others; for Sharp/10%, which exhibited the poorest visual similarity, PSNR increased from a mean ± CI of 17.763 ± 0.492 to 31.925 ± 0.571, SSIM from 0.219 ± 0.009 to 0.754 ± 0.017, and LPIPS decreased from 0.490 ± 0.005 to 0.275 ± 0.016. Texture-based radiomic features exhibited a greater degree of variability across conditions, i.e. a CCC of 0.500 ± 0.332, compared to intensity-based features (0.972 ± 0.045). GANs achieved the highest CCC (0.969 ± 0.009 for radiomic and 0.841 ± 0.070 for deep features) amongst others. CNNs are suitable if downstream applications necessitate visual interpretation of images, whereas GANs are better alternatives for generating reproducible quantitative image features needed for machine learning applications.Significance. Understanding the efficacy of harmonization in addressing multi-parameter variability is crucial for optimizing diagnostic accuracy and a critical step toward building generalizable models suitable for clinical use.

Developing a novel dosiomics model to predict treatment failures following lung stereotactic body radiation therapy

Purpose

The purpose of this study was to investigate the dosiomics features of the interplay between CT density and dose distribution in lung SBRT plans, and to develop a model to predict treatment failure following lung SBRT treatment.

Methods

A retrospective study was conducted involving 179 lung cancer patients treated with SBRT at the University of Nebraska Medical Center (UNMC) between October 2007 and June 2022. Features from the CT image, Biological Effective Dose (BED) and five interaction matrices between CT and BED were extracted using radiomics mathematics. Our in-house feature selection pipeline was utilized to evaluate and rank features based on their importance and redundancy, with only the selected non-redundant features being used for predictive modeling. We randomly selected 151 cases and 28 cases as training and test datasets. Four different models were trained utilizing the Balanced Random Forest framework on the same training dataset to differentiate between failure and non-failure cases. These four models utilized the same number of selected features extracted from CT-only, BED-only, a combination of CT and BED, and a composite of CT and BED including their interaction matrices, respectively.

Results

The cohort included 125 non-failure cases and 54 failure cases, with a median follow-up time of 34.4 months. We selected the top 17 important and non-redundant features (with the Pearsons's coefficient < 0.5) in each model. When evaluated on the same independent test set, the four models-CT features-only, BED features-only, a combination of CT and BED features, and a composite model including features from CT and BED that includes their interaction matrices-achieved AUC values of 0.56, 0.75, 0.73, and 0.82, respectively, with corresponding accuracies of 0.61, 0.79, 0.71, and 0.79. The composite model demonstrated the highest AUC and accuracy, indicating that incorporating interactions between CT and BED reveals more predictive capabilities in distinguishing between failure and non-failure cases.

Conclusion

The dosiomics model integrating the interaction between CT and Dose can effectively predict treatment failure following lung SBRT treatment and may serve as a useful tool to proactively evaluate and select lung SBRT treatment plans to reduce treatment failure in the future.

Annotated test-retest dataset of lung cancer CT scan images reconstructed at multiple imaging parameters

Quantitative imaging biomarkers (QIB) are increasingly used in clinical research to advance precision medicine approaches in oncology. Computed tomography (CT) is a modality of choice for cancer diagnosis, prognosis, and response assessment due to its reliability and global accessibility. Here, we contribute to the cancer imaging community through The Cancer Imaging Archive (TCIA) by providing investigator-initiated, same-day repeat CT scan images of 32 non-small cell lung cancer (NSCLC) patients, along with radiologist-annotated lesion contours as a reference standard. Each scan was reconstructed into 6 image settings using various combinations of three slice thicknesses (1.25 mm, 2.5 mm, 5 mm) and two reconstruction kernels (lung, standard; GE CT equipment), which spans a wide range of CT imaging reconstruction parameters commonly used in lung cancer clinical practice and clinical trials. This holds considerable value for advancing the development of robust Radiomics, Artificial Intelligence (AI) and machine learning (ML) methods.

Predicting EGFR Mutation Status in Non-Small Cell Lung Cancer Using Artificial Intelligence: A Systematic Review and Meta-Analysis

RATIONALE AND OBJECTIVES

Recent advancements in artificial intelligence (AI) render a substantial promise for epidermal growth factor receptor (EGFR) mutation status prediction in non-small cell lung cancer (NSCLC). We aimed to evaluate the performance and quality of AI algorithms that use radiomics features in predicting EGFR mutation status in patient with NSCLC.

MATERIALS AND METHODS

We searched PubMed (Medline), EMBASE, Web of Science, and IEEExplore for studies published up to February 28, 2022. Studies utilizing an AI algorithm (either conventional machine learning [cML] and deep learning [DL]) for predicting EGFR mutations in patients with NSLCL were included. We extracted binary diagnostic accuracy data and constructed a bivariate random-effects model to obtain pooled sensitivity, specificity, and 95% confidence interval. This study is registered with PROSPERO, CRD42021278738.

RESULTS

Our search identified 460 studies, of which 42 were included. Thirty-five studies were included in the meta-analysis. The AI algorithms exhibited an overall area under the curve (AUC) value of 0.789 and pooled sensitivity and specificity levels of 72.2% and 73.3%, respectively. The DL algorithms outperformed cML in terms of AUC (0.822 vs. 0.775) and sensitivity (80.1% vs. 71.1%), but had lower specificity (70.0% vs. 73.8%, p-value < 0.001) compared to cML. Subgroup analysis revealed that the use of positron-emission tomography/computed tomography, additional clinical information, deep feature extraction, and manual segmentation can improve diagnostic performance.

CONCLUSION

DL algorithms can serve as a novel method for increasing predictive accuracy and thus have considerable potential for use in predicting EGFR mutation status in patient with NSCLC. We also suggest that guidelines on using AI algorithms in medical image analysis should be developed with a focus on oncologic radiomics.

Radiomics of metastatic brain tumor as a predictive image biomarker of progression-free survival in patients with non-small-cell lung cancer with brain metastasis receiving tyrosine kinase inhibitors

BACKGROUND AND OBJECTIVE

Epidermal growth factor receptor (EGFR)-targeted tyrosine kinase inhibitors (TKIs) are the first-line therapy for EGFR-mutant non-small-cell lung cancer (NSCLC). Early prediction of treatment failure in patients with brain metastases treated with EGFR-TKIs may help in making decisions for systemic drug therapy or local brain tumor control. This study examined the predictive power of the radiomics of both brain metastasis tumors and primary lung tumors. We propose a deep learning based CoxCC model based on quantitative brain magnetic resonance imaging (MRI), a prognostic index and clinical data; the model can be used to predict progression-free survival (PFS) after EGFR-TKI therapy in advanced EGFR-mutant NSCLC.

METHODS

This retrospective single-center study included 271 patients receiving first-line EGFR-TKI targeted therapy in 2018-2019. Among them, 72 patients who had brain metastases before receiving first-line EGFR-TKI treatment. Three radiomic features were extracted from pretreatment brain MRI images. A CoxCC model for the progression risk stratification of EGFR-TKI treatment was proposed on the basis of MRI radiomics, clinical features, and a prognostic index. We performed time-dependent PFS predictions to evaluate the performance of the CoxCC model.

RESULTS

The CoxCC model based on a prognostic index, clinical features, and radiomic features of brain metastasis exhibited higher performance than clinical features combined with indexes previously proposed for determining the prognosis of brain metastasis, including recursive partitioning analysis, diagnostic-specific graded prognostic assessment, graded prognostic assessment for lung cancer using molecular markers (lung-molGPA), and modified lung-molGPA, with c-index values of 0.75, 0.67, 0.66, 0.65, and 0.65, respectively. The model achieved areas under the curve of 0.88, 0.73, 0.92, and 0.90 for predicting PFS at 3, 6, 9 and 12 months, respectively. PFS significantly differed between the high- and low-risk groups (p < 0.001).

CONCLUSIONS

For patients with advanced-stage NSCLC with brain metastasis, MRI radiomics of brain metastases may predict PFS. The CoxCC model integrating brain metastasis radiomics, clinical features, and a prognostic index provided reliable multi-time-point PFS predictions for patients with advanced NSCLC and brain metastases receiving EGFR-TKI treatment.

Current status and prospect of PET-related imaging radiomics in lung cancer

Lung cancer is highly aggressive, which has a high mortality rate. Major types encompass lung adenocarcinoma, lung squamous cell carcinoma, lung adenosquamous carcinoma, small cell carcinoma, and large cell carcinoma. Lung adenocarcinoma and lung squamous cell carcinoma together account for more than 80% of cases. Diverse subtypes demand distinct treatment approaches. The application of precision medicine necessitates prompt and accurate evaluation of treatment effectiveness, contributing to the improvement of treatment strategies and outcomes. Medical imaging is crucial in the diagnosis and management of lung cancer, with techniques such as fluoroscopy, computed radiography (CR), digital radiography (DR), computed tomography (CT), magnetic resonance imaging (MRI), positron emission tomography (PET)/CT, and PET/MRI being essential tools. The surge of radiomics in recent times offers fresh promise for cancer diagnosis and treatment. In particular, PET/CT and PET/MRI radiomics, extensively studied in lung cancer research, have made advancements in diagnosing the disease, evaluating metastasis, predicting molecular subtypes, and forecasting patient prognosis. While conventional imaging methods continue to play a primary role in diagnosis and assessment, PET/CT and PET/MRI radiomics simultaneously provide detailed morphological and functional information. This has significant clinical potential value, offering advantages for lung cancer diagnosis and treatment. Hence, this manuscript provides a review of the latest developments in PET-related radiomics for lung cancer.

AI/ML advances in non-small cell lung cancer biomarker discovery

Lung cancer is the leading cause of cancer deaths among both men and women, representing approximately 25% of cancer fatalities each year. The treatment landscape for non-small cell lung cancer (NSCLC) is rapidly evolving due to the progress made in biomarker-driven targeted therapies. While advancements in targeted treatments have improved survival rates for NSCLC patients with actionable biomarkers, long-term survival remains low, with an overall 5-year relative survival rate below 20%. Artificial intelligence/machine learning (AI/ML) algorithms have shown promise in biomarker discovery, yet NSCLC-specific studies capturing the clinical challenges targeted and emerging patterns identified using AI/ML approaches are lacking. Here, we employed a text-mining approach and identified 215 studies that reported potential biomarkers of NSCLC using AI/ML algorithms. We catalogued these studies with respect to BEST (Biomarkers, EndpointS, and other Tools) biomarker sub-types and summarized emerging patterns and trends in AI/ML-driven NSCLC biomarker discovery. We anticipate that our comprehensive review will contribute to the current understanding of AI/ML advances in NSCLC biomarker research and provide an important catalogue that may facilitate clinical adoption of AI/ML-derived biomarkers.

Evaluating the Potential of Delta Radiomics for Assessing Tyrosine Kinase Inhibitor Treatment Response in Non-Small Cell Lung Cancer Patients

Our study aimed to harness the power of CT scans, observed over time, in predicting how lung adenocarcinoma patients might respond to a treatment known as EGFR-TKI. Analyzing scans from 322 advanced stage lung cancer patients, we identified distinct image-based patterns. By integrating these patterns with comprehensive clinical information, such as gene mutations and treatment regimens, our predictive capabilities were significantly enhanced. Interestingly, the precision of these predictions, particularly related to radiomics features, diminished when data from various centers were combined, suggesting that the approach requires standardization across facilities. This novel method offers a potential pathway to anticipate disease progression in lung adenocarcinoma patients treated with EGFR-TKI, laying the groundwork for more personalized treatments. To further validate this approach, extensive studies involving a larger cohort are pivotal.

Systematic Review, Meta-Analysis and Radiomics Quality Score Assessment of CT Radiomics-Based Models Predicting Tumor EGFR Mutation Status in Patients with Non-Small-Cell Lung Cancer

Assessment of the quality and current performance of computed tomography (CT) radiomics-based models in predicting epidermal growth factor receptor (EGFR) mutation status in patients with non-small-cell lung carcinoma (NSCLC). Two medical literature databases were systematically searched, and articles presenting original studies on CT radiomics-based models for predicting EGFR mutation status were retrieved. Forest plots and related statistical tests were performed to summarize the model performance and inter-study heterogeneity. The methodological quality of the selected studies was assessed via the Radiomics Quality Score (RQS). The performance of the models was evaluated using the area under the curve (ROC AUC). The range of the Risk RQS across the selected articles varied from 11 to 24, indicating a notable heterogeneity in the quality and methodology of the included studies. The average score was 15.25, which accounted for 42.34% of the maximum possible score. The pooled Area Under the Curve (AUC) value was 0.801, indicating the accuracy of CT radiomics-based models in predicting the EGFR mutation status. CT radiomics-based models show promising results as non-invasive alternatives for predicting EGFR mutation status in NSCLC patients. However, the quality of the studies using CT radiomics-based models varies widely, and further harmonization and prospective validation are needed before the generalization of these models.

uRP: An integrated research platform for one-stop analysis of medical images

Introduction

Medical image analysis is of tremendous importance in serving clinical diagnosis, treatment planning, as well as prognosis assessment. However, the image analysis process usually involves multiple modality-specific software and relies on rigorous manual operations, which is time-consuming and potentially low reproducible.

Methods

We present an integrated platform - uAI Research Portal (uRP), to achieve one-stop analyses of multimodal images such as CT, MRI, and PET for clinical research applications. The proposed uRP adopts a modularized architecture to be multifunctional, extensible, and customizable.

Results and Discussion

The uRP shows 3 advantages, as it 1) spans a wealth of algorithms for image processing including semi-automatic delineation, automatic segmentation, registration, classification, quantitative analysis, and image visualization, to realize a one-stop analytic pipeline, 2) integrates a variety of functional modules, which can be directly applied, combined, or customized for specific application domains, such as brain, pneumonia, and knee joint analyses, 3) enables full-stack analysis of one disease, including diagnosis, treatment planning, and prognosis assessment, as well as full-spectrum coverage for multiple disease applications. With the continuous development and inclusion of advanced algorithms, we expect this platform to largely simplify the clinical scientific research process and promote more and better discoveries.

How scan parameter choice affects deep learning-based coronary artery disease assessment from computed tomography

Recently, algorithms capable of assessing the severity of Coronary Artery Disease (CAD) in form of the Coronary Artery Disease-Reporting and Data System (CAD-RADS) grade from Coronary Computed Tomography Angiography (CCTA) scans using Deep Learning (DL) were proposed. Before considering to apply these algorithms in clinical practice, their robustness regarding different commonly used Computed Tomography (CT)-specific image formation parameters-including denoising strength, slab combination, and reconstruction kernel-needs to be evaluated. For this study, we reconstructed a data set of 500 patient CCTA scans under seven image formation parameter configurations. We select one default configuration and evaluate how varying individual parameters impacts the performance and stability of a typical algorithm for automated CAD assessment from CCTA. This algorithm consists of multiple preprocessing and a DL prediction step. We evaluate the influence of the parameter changes on the entire pipeline and additionally on only the DL step by propagating the centerline extraction results of the default configuration to all others. We consider the standard deviation of the CAD severity prediction grade difference between the default and variation configurations to assess the stability w.r.t. parameter changes. For the full pipeline we observe slight instability (± 0.226 CAD-RADS) for all variations. Predictions are more stable with centerlines propagated from the default to the variation configurations (± 0.122 CAD-RADS), especially for differing denoising strengths (± 0.046 CAD-RADS). However, stacking slabs with sharp boundaries instead of mixing slabs in overlapping regions (called true stack ± 0.313 CAD-RADS) and increasing the sharpness of the reconstruction kernel (± 0.150 CAD-RADS) leads to unstable predictions. Regarding the clinically relevant tasks of excluding CAD (called rule-out; AUC default 0.957, min 0.937) and excluding obstructive CAD (called hold-out; AUC default 0.971, min 0.964) the performance remains on a high level for all variations. Concluding, an influence of reconstruction parameters on the predictions is observed. Especially, scans reconstructed with the true stack parameter need to be treated with caution when using a DL-based method. Also, reconstruction kernels which are underrepresented in the training data increase the prediction uncertainty.

Relationship between the image characteristics of artificial intelligence and EGFR gene mutation in lung adenocarcinoma

Lung Adenocarcinoma (LUAD) is a kind of Lung Cancer (LCA) with high incidence rate, which is very harmful to human body. It is hidden in the human body and is not easy to be discovered, so it brings great inconvenience to the treatment of LUAD. Artificial Intelligence (AI) technology provides technical support for the diagnosis and treatment of LUAD and has great application space in intelligent medicine. In this paper, 164 patients with primary LUAD who underwent surgery in Hospital A from January 2020 to December 2021 were selected as the study subjects, and the correlation between the imaging characteristics of LUAD and Epidermal Growth Factor Receptor (EGFR) gene mutation was analyzed. Finally, the conclusion was drawn. In terms of the study on the correlation between EGFR mutation of LUAD and the imaging characteristics of Computed Tomography (CT), it was concluded that there were significant differences between the patient's sex, smoking history, pulmonary nodule morphology and the EGFR gene, and there was no significant difference between the patient's tumor size and EGFR gene; in the study of the relationship between EGFR gene mutation and CT signs of LUAD lesions, it was found that there were significant differences between the symptoms of cavity sign, hair prick sign and chest depression sign and EGFR gene, but there was no significant difference between the symptoms of lobulation sign and EGFR gene; in the study of pathological subtype and EGFR gene mutation status of LUAD patients, it was concluded that the pathological subtype was mainly micropapillary. The mutation rate was 44.44%, which was the highest; in terms of CT manifestations of adjacent structures of lung cancer and the study of EGFR gene mutation status, it was found that there was a statistical difference between the tumor with vascular convergence sign and EGFR gene mutation, and pleural effusion, pericardial effusion, pleural thickening and other signs in tumor imaging were not significantly associated with EGFR gene mutation; in terms of the study of CT manifestations of adjacent structures of LCA and EGFR gene mutation status, it was concluded that pleural effusion, pericardial effusion, pleural thickening and other signs in tumor images were not significantly associated with EGFR gene mutation; in terms of analysis and cure of LUAD, it was concluded that the cure rate of patients was relatively high, and only a few people died of ineffective treatment. This paper provided a reference for the field of intelligent medicine and physical health.

Preoperative prediction of mediastinal lymph node metastasis in non-small cell lung cancer based on 18F-FDG PET/CT radiomics

AIM

To establish and verify a 2-[¹⁸F]-fluoro-2-deoxy-d-glucose (FDG) positron-emission tomography (PET)/computed tomography (CT)-based radiomics nomogram to predict mediastinal lymph node metastasis (LNM) in non-small cell lung cancer (NSCLC) patients preoperatively.

MATERIALS AND METHODS

This retrospective study enrolled 155 NSCLC patients (primary cohort, n=93; validation cohort, n=62). For each patient, 2,704 radiomic features were extracted from the primary lung cancer regions. Four procedures including the Mann-Whitney U-test, Spearman's correlation analysis, minimum redundancy-maximum relevance (mRMR), and least absolute shrinkage and selection operator (LASSO) binary logistic regression were utilised for determining essential features and establishing a radiomics signature. After that, a nomogram was established. The nomogram's potential was assessed based on its discrimination, calibration, and clinical usefulness. The radiomics signature and nomogram predictive performances were evaluated with respect to the area under the receiver operating characteristic curve (AUC), specificity, accuracy, and sensitivity.

RESULTS

The radiomics signature composed of eight selected features had good discriminatory performance of LNM versus non-LNM groups an AUC of 0.851 and 0.826 in primary and validation cohorts, respectively. The nomogram also indicated good discrimination with an AUC of 0.869 and 0.847 in the primary and validation cohorts, respectively. Furthermore, good calibration was demonstrated utilising the nomogram.

CONCLUSIONS

An ¹⁸F-FDG PET/CT-based radiomics nomogram that integrates the radiomics signature and age was promoted to predict mediastinal LNM within NSCLC patients, which could potentially facilitate individualised therapy for mediastinal LNM before treatment. The nomogram was beneficial in clinical practice, as illustrated by decision curve analysis.

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.

Prognostic analysis and risk stratification of lung adenocarcinoma undergoing EGFR-TKI therapy with time-serial CT-based radiomics signature

OBJECTIVES

To evaluate the value of time-serial CT radiomics features in predicting progression-free survival (PFS) for lung adenocarcinoma (LUAD) patients after epidermal growth factor receptor-tyrosine kinase inhibitors (EGFR-TKIs) therapy.

MATERIALS AND METHODS

LUAD patients treated with EGFR-TKIs were retrospectively included from three independent institutes and divided into training and validation cohorts. Intratumoral and peritumoral features were extracted from time-serial non-contrast chest CT (including pre-therapy and first follow-up images); moreover, the percentage variation per unit time (day) was introduced to adjust for the different follow-up periods of each patient. Test-retest was performed to exclude irreproducible features, while the Boruta algorithm was used to select critical radiomics features. Radiomics signatures were constructed with random forest survival models in the training cohort and compared against baseline clinical characteristics through Cox regression and nonparametric testing of concordance indices (C-indices).

RESULTS

The training cohort included 131 patients (74 women, 56.5%) from one institute and the validation cohort encompassed 41 patients (24 women, 58.5%) from two other institutes. The optimal signature contained 10 features and 7 were unit time feature variations. The comprehensive radiomics model outperformed the pre-therapy clinical characteristics in predicting PFS (training: 0.78, 95% CI: [0.72, 0.84] versus 0.55, 95% CI: [0.49, 0.62], p < 0.001; validation: 0.72, 95% CI: [0.60, 0.84] versus 0.54, 95% CI: [0.42, 0.66], p < 0.001).

CONCLUSION

Radiomics signature derived from time-serial CT images demonstrated optimal prognostic performance of disease progression. This dynamic imaging biomarker holds the promise of monitoring treatment response and achieving personalized management.

KEY POINTS

• The intrinsic tumor heterogeneity can be highly dynamic under the therapeutic effect of EGFR-TKI treatment, and the inevitable development of drug resistance may disrupt the duration of clinical benefit. Decision-making remained challenging in practice to detect the emergence of acquired resistance during the early response phase. • Time-serial CT-based radiomics signature integrating intra- and peritumoral features offered the potential to predict progression-free survival for LUAD patients treated with EGFR-TKIs. • The dynamic imaging signature allowed for prognostic risk stratification.

Can Persistent Homology Features Capture More Intrinsic Information about Tumors from 18F-Fluorodeoxyglucose Positron Emission Tomography/Computed Tomography Images of Head and Neck Cancer Patients?

This study hypothesized that persistent homology (PH) features could capture more intrinsic information about the metabolism and morphology of tumors from ¹⁸F-fluorodeoxyglucose positron emission tomography (PET)/computed tomography (CT) images of patients with head and neck (HN) cancer than other conventional features. PET/CT images and clinical variables of 207 patients were selected from the publicly available dataset of the Cancer Imaging Archive. PH images were generated from persistent diagrams obtained from PET/CT images. The PH features were derived from the PH PET/CT images. The signatures were constructed in a training cohort from features from CT, PET, PH-CT, and PH-PET images; clinical variables; and the combination of features and clinical variables. Signatures were evaluated using statistically significant differences (p-value, log-rank test) between survival curves for low- and high-risk groups and the C-index. In an independent test cohort, the signature consisting of PH-PET features and clinical variables exhibited the lowest log-rank p-value of 3.30 × 10⁻⁵ and C-index of 0.80, compared with log-rank p-values from 3.52 × 10⁻² to 1.15 × 10⁻⁴ and C-indices from 0.34 to 0.79 for other signatures. This result suggests that PH features can capture the intrinsic information of tumors and predict prognosis in patients with HN cancer.

Artificial intelligence and radiomics: fundamentals, applications, and challenges in immunotherapy

Immunotherapy offers the potential for durable clinical benefit but calls into question the association between tumor size and outcome that currently forms the basis for imaging-guided treatment. Artificial intelligence (AI) and radiomics allow for discovery of novel patterns in medical images that can increase radiology’s role in management of patients with cancer, although methodological issues in the literature limit its clinical application. Using keywords related to immunotherapy and radiomics, we performed a literature review of MEDLINE, CENTRAL, and Embase from database inception through February 2022. We removed all duplicates, non-English language reports, abstracts, reviews, editorials, perspectives, case reports, book chapters, and non-relevant studies. From the remaining articles, the following information was extracted: publication information, sample size, primary tumor site, imaging modality, primary and secondary study objectives, data collection strategy (retrospective vs prospective, single center vs multicenter), radiomic signature validation strategy, signature performance, and metrics for calculation of a Radiomics Quality Score (RQS). We identified 351 studies, of which 87 were unique reports relevant to our research question. The median (IQR) of cohort sizes was 101 (57–180). Primary stated goals for radiomics model development were prognostication (n=29, 33.3%), treatment response prediction (n=24, 27.6%), and characterization of tumor phenotype (n=14, 16.1%) or immune environment (n=13, 14.9%). Most studies were retrospective (n=75, 86.2%) and recruited patients from a single center (n=57, 65.5%). For studies with available information on model testing, most (n=54, 65.9%) used a validation set or better. Performance metrics were generally highest for radiomics signatures predicting treatment response or tumor phenotype, as opposed to immune environment and overall prognosis. Out of a possible maximum of 36 points, the median (IQR) of RQS was 12 (10–16). While a rapidly increasing number of promising results offer proof of concept that AI and radiomics could drive precision medicine approaches for a wide range of indications, standardizing the data collection as well as optimizing the methodological quality and rigor are necessary before these results can be translated into clinical practice.

Differentiating EGFR from ALK mutation status using radiomics signature based on MR sequences of brain metastasis

PURPOSE

More and more small brain metastases (BMs) in asymptomatic patients can be detected even prior to their primary lung cancer with the development of MRI. The aim of this study was to develop a predictive radiomics model to identify epidermal growth factor receptor (EGFR) and anaplastic lymphoma kinase (ALK) mutation status in BM and explore the optimal MR sequence for predication.

METHODS

This retrospective study included 186 patients with proven BM of lung cancer (training cohort: 70 patients with EGFR mutations and 65 patients with ALK rearrangements; testing cohort: 26 patients with EGFR mutations and 25 patients with ALK rearrangements). Radiomics features were separately extracted from contrast-enhanced T1-weighted imaging (T1-CE), T2 fluid-attenuated inversion recovery (T2-FLAIR) and T2WI sequences. The model for three MR sequences were constructed using a random forest classifier. ROC curves were used to validate the capability of the models in the training and testing cohorts.

RESULTS

The AUCs of the T2-FLAIR model were significantly higher than those of the T1-CE model in training cohort (0.991 versus 0.954) and testing cohort (0.950 versus 0.867) and much higher than those of the T2WI model in training cohort (0.991 versus 0.880) and testing cohort (0.950 versus 0.731). Besides, the F1 scores of the T1-CE model were slightly higher than the T2-FLAIR model and much higher than the T2WI model in two cohorts.

CONCLUSION

T2-FLAIR and T1-CE radiomics models that can be used as noninvasive tools for identifying EGFR and ALK mutation status are helpful to guide therapeutic strategies.

Assessing Outcomes in NSCLC: Radiomic analysis, kinetic analysis and circulating tumor DNA

Current radiographic methods of measuring treatment response for patients with nonsmall cell lung cancer have significant limitations. Recently, new modalities using standard of care images or minimally invasive blood-based DNA tests have gained interest as methods of evaluating treatment response. This article highlights three emerging modalities: radiomic analysis, kinetic analysis and serum-based measurement of circulating tumor DNA, with a focus on the clinical evidence supporting these methods. Additionally, we discuss the possibility of combining these modalities to develop a robust biomarker with strong correlation to clinically meaningful outcomes that could impact clinical trial design and patient care. At Last, we focus on how these methods specifically apply to a Veteran population.

Different CT slice thickness and contrast-enhancement phase in radiomics models on the differential performance of lung adenocarcinoma

Background

To investigate the effects of computed tomography (CT) reconstruction slice thickness and contrast‐enhancement phase on the differential diagnosis performance of radiomic signature in lung adenocarcinoma.

Methods

A total of 187 patients who had been pathologically confirmed with lung adenocarcinoma and nonadenocarcinoma were divided into a training cohort (n = 149) and validation cohort (n = 38). All the patients underwent contrast‐enhanced CT and the images were reconstructed with different slice thickness. The radiomic features were extracted from different slice thickness and scan phase. The logistic regression (LR) algorithm was used to build a machine learning model for each group. The area under the curve (AUC) obtained from the receiver operating characteristic (ROC) curve and DeLong test was used to evaluate its discriminating performance.

Results

Finally, 34 image features and five semantic features were selected to establish a radiomics model. Based on the three contrast‐enhanced CT phases and four reconstruction slice thickness, 12 groups of radiomics models showed good discrimination ability with the AUCs range from 0.9287 to 0.9631, sensitivity range from 0.8349 to 0.9083, specificity range from 0.825 to 0.925 in the training group. Similar results were observed in the validation group. However, there was no statistical significance between the different CT scan phase groups and different slice thickness (p > 0.05).

Conclusions

The radiomic analysis of contrast‐enhanced CT can be used for the differential diagnosis of lung adenocarcinoma. Moreover, different slice thickness and contrast‐enhanced scan phase did not affect the discriminating ability in the radiomics models.

Molecular typing of lung adenocarcinoma with computed tomography and CT image-based radiomics: a narrative review of research progress and prospects

Objective

The purpose of this paper was to perform a narrative review of current research evidence on conventional computed tomography (CT) imaging features and CT image-based radiomic features for predicting gene mutations in lung adenocarcinoma and discuss how to translate the research findings to guide future practice.

Background

Lung cancer, especially lung adenocarcinoma, is the leading cause of cancer-related deaths. With advances in the diagnosis and treatment of lung adenocarcinoma with the emergence of molecular testing, the prediction of oncogenes and even drug resistance gene mutations have become key to individualized and precise clinical treatment in order to prolong survival and improve quality of life. The progress of imageological examination includes the development of CT and radiomics are promising quantitative methods for predicting different gene mutations in lung adenocarcinoma, especially common mutations, such as epidermal growth factor receptor (EGFR) mutation, anaplastic lymphoma kinase (ALK) mutation and Kirsten rat sarcoma viral oncogene (KRAS) mutation.

Methods

The PubMed electronic database was searched along with a set of terms specific to lung adenocarcinoma, radiomics (including texture analysis), CT, computed tomography, EGFR, ALK, KRAS, rearranging transfection (RET) rearrangement and c-ros oncogene 1 (ROS-1), v-raf murine sarcoma viral oncogene homolog B1 (BRAF), and human epidermal growth factor receptor 2 (HER2) mutations et al. This review has been reported in compliance with the Narrative Review checklist guidelines. From each full-text article, information was extracted regarding a set of terms above.

Conclusions

Research on the application of conventional CT features and CT image-based radiomic features for predicting the gene mutation status of lung adenocarcinoma is still in a preliminary stage. Noninvasively determination of mutation status in lung adenocarcinoma before targeted therapy with conventional CT features and CT image-based radiomic features remains both hopes and challenges. Before radiomics could be applied in clinical practice, more work needs to be done.

Maximum Standardized Uptake Value of 18F-deoxyglucose PET Imaging Increases the Effectiveness of CT Radiomics in Differentiating Benign and Malignant Pulmonary Ground-Glass Nodules

To investigate whether the maximum standardized uptake value (SUVmax) of ¹⁸F-deoxyglucose (FDG) PET imaging can increase the diagnostic efficiency of CT radiomics-based prediction model in differentiating benign and malignant pulmonary ground-glass nodules (GGNs). We retrospectively collected 190 GGNs from 165 patients who underwent ¹⁸F-FDG PET/CT examination from January 2012 to March 2020. Propensity score matching (PSM) was performed to select GGNs with similar baseline characteristics. LIFEx software was used to extract 49 CT radiomic features, and the least absolute shrinkage and selection operator (LASSO) algorithm was used to select parameters and establish the Rad-score. Logistic regression analysis was performed combined with semantic features to construct a CT radiomics model, which was combined with SUVmax to establish the PET + CT radiomics model. Receiver operating characteristic (ROC) was used to compare the diagnostic efficacy of different models. After PSM at 1:4, 190 GGNs were divided into benign group (n = 23) and adenocarcinoma group (n = 92). After texture analysis, the Rad-score with three CT texture features was constructed for each nodule. Compared with the Rad-score and CT radiomics model (AUC: 0.704 (95%CI: 0.562-0.845) and 0.908 (95%CI: 0.842-0.975), respectively), PET + CT radiomics model had the best diagnostic efficiency (AUC: 0.940, 95%CI: 0.889-0.990), and there was significant difference between each two of them (P = 0.001-0.030). SUVmax can effectively improve CT radiomics model performance in the differential diagnosis of benign and malignant GGNs. PET + CT radiomics might become a noninvasive and reliable method for differentiating of GGNs.

Convolutional Neural Network Addresses the Confounding Impact of CT Reconstruction Kernels on Radiomics Studies

Achieving high feature reproducibility while preserving biological information is one of the main challenges for the generalizability of current radiomics studies. Non-clinical imaging variables, such as reconstruction kernels, have shown to significantly impact radiomics features. In this study, we retrain an open-source convolutional neural network (CNN) to harmonize computerized tomography (CT) images with various reconstruction kernels to improve feature reproducibility and radiomic model performance using epidermal growth factor receptor (EGFR) mutation prediction in lung cancer as a paradigm. In the training phase, the CNN was retrained and tested on 32 lung cancer patients’ CT images between two different groups of reconstruction kernels (smooth and sharp). In the validation phase, the retrained CNN was validated on an external cohort of 223 lung cancer patients’ CT images acquired using different CT scanners and kernels. The results showed that the retrained CNN could be successfully applied to external datasets with different CT scanner parameters, and harmonization of reconstruction kernels from sharp to smooth could significantly improve the performance of radiomics model in predicting EGFR mutation status in lung cancer. In conclusion, the CNN based method showed great potential in improving feature reproducibility and generalizability by harmonizing medical images with heterogeneous reconstruction kernels.

A Combination Model of Radiomics Features and Clinical Biomarkers as a Nomogram to Differentiate Nonadvanced From Advanced Liver Fibrosis: A Retrospective Study

RATIONALE AND OBJECTIVES

To develop and validate a combination model of radiomics features and clinical biomarkers to differentiate nonadvanced from advanced liver fibrosis.

MATERIALS AND METHODS

One hundred and eight consecutive patients with pathologically diagnosed liver fibrosis were randomly placed in a training or a test cohort at a ratio of 2:1. For each patient, 1674 radiomics features extracted from portal venous phase CT images were reduced by using minimum redundancy and maximum relevant. The optimal features identified were incorporated into the radiomics model. Eight clinical markers were evaluated. Integrated with clinical independent risk factors, a combination model was built. A nomogram was also established from the model. The performance of the models was assessed. Finally, a decision curve analysis was performed to estimate the clinical usefulness of the nomogram.

RESULTS

The radiomics model established using five features achieved a promising level of discrimination between nonadvanced and advanced liver fibrosis. The combination model incorporated the radiomics signature with two clinical biomarkers and showed good calibration and discrimination. The training and testing cohort results of the radiomics model were area under curve values 0.864 and 0.772, accuracy 77.8% and 77.8%, sensitivity 86.7% and 73.1%, and specificity 71.4% and 90.0%, respectively. For the combination model, the training and testing cohort results were area under curve values 0.915 and 0.897, accuracy 83.3% and 86.1%, sensitivity 86% and 80.6%, and specificity 82.6% and 92.3%, respectively. The decision curve indicated the nomogram has potential in clinical application.

CONCLUSION

This combination model provides a promising approach for differentiating non-advanced from advanced liver fibrosis.

Predicting Tumor Budding Status in Cervical Cancer Using MRI Radiomics: Linking Imaging Biomarkers to Histologic Characteristics

BACKGROUND

Our previous study demonstrated that tumor budding (TB) status was associated with inferior overall survival in cervical cancer. The purpose of this study is to evaluate whether radiomic features can predict TB status in cervical cancer patients.

METHODS

Seventy-four patients with cervical cancer who underwent preoperative MRI and radical hysterectomy from 2011 to 2015 at our institution were enrolled. The patients were randomly allocated to the training dataset (n = 48) and test dataset (n = 26). Tumors were segmented on axial gadolinium-enhanced T1- and T2-weighted images. A total of 2074 radiomic features were extracted. Four machine learning classifiers, including logistic regression (LR), random forest (RF), support vector machine (SVM), and neural network (NN), were used. The trained models were validated on the test dataset.

RESULTS

Twenty radiomic features were selected; all were features from filtered-images and 85% were texture-related features. The area under the curve values and accuracy of the models by LR, RF, SVM and NN were 0.742 and 0.769, 0.782 and 0.731, 0.849 and 0.885, and 0.891 and 0.731, respectively, in the test dataset.

CONCLUSION

MRI-based radiomic features could predict TB status in patients with cervical cancer.

Effect of CT image acquisition parameters on diagnostic performance of radiomics in predicting malignancy of pulmonary nodules of different sizes

OBJECTIVES

To investigate the effect of CT image acquisition parameters on the performance of radiomics in classifying benign and malignant pulmonary nodules (PNs) with respect to nodule size.

METHODS

We retrospectively collected CT images of 696 patients with PNs from March 2015 to March 2018. PNs were grouped by nodule diameter: T1a (diameter ≤ 1.0 cm), T1b (1.0 cm < diameter ≤ 2.0 cm), and T1c (2.0 cm < diameter ≤ 3.0 cm). CT images were divided into four settings according to slice-thickness-convolution-kernels: setting 1 (slice thickness/reconstruction type: 1.25 mm sharp), setting 2 (5 mm sharp), setting 3 (5 mm smooth), and random setting. We created twelve groups from two interacting conditions. Each PN was segmented and had 1160 radiomics features extracted. Non-redundant features with high predictive ability in training were selected to build a distinct model under each of the twelve subsets.

RESULTS

The performance (AUCs) on predicting PN malignancy were as follows: T1a group: 0.84, 0.64, 0.68, and 0.68; T1b group: 0.68, 0.74, 0.76, and 0.70; T1c group: 0.66, 0.64, 0.63, and 0.70, for the setting 1, setting 2, setting 3, and random setting, respectively. In the T1a group, the AUC of radiomics model in setting 1 was statistically significantly higher than all others; In the T1b group, AUCs of radiomics models in setting 3 were statistically significantly higher than some; and in the T1c group, there were no statistically significant differences among models.

CONCLUSIONS

For PNs less than 1 cm, CT image acquisition parameters have a significant influence on diagnostic performance of radiomics in predicting malignancy, and a model created using images reconstructed with thin section and a sharp kernel algorithm achieved the best performance. For PNs larger than 1 cm, CT reconstruction parameters did not affect diagnostic performance substantially.

KEY POINTS

• CT image acquisition parameters have a significant influence on the diagnostic performance of radiomics in pulmonary nodules less than 1 cm. • In pulmonary nodules less than 1 cm, a radiomics model created by using images reconstructed with thin section and a sharp kernel algorithm achieved the best diagnostic performance. • For PNs larger than 1 cm, CT image acquisition parameters do not affect diagnostic performance substantially.

Deep neural network-based approach to improving radiomics analysis reproducibility in liver cancer: effect on image resampling

Objectives.To test the effect of traditional up-sampling slice thickness (ST) methods on the reproducibility of CT radiomics features of liver tumors and investigate the improvement using a deep neural network (DNN) scheme.Methods.CT images with ≤ 1 mm ST in the public dataset were converted to low-resolution (3 mm, 5 mm) CT images. A DNN model was trained for the conversion from 3 mm ST and 5 mm ST to 1 mm ST and compared with conventional interpolation-based methods (cubic, linear, nearest) using structural similarity (SSIM) and peak-signal-to-noise-ratio (PSNR). Radiomics features were extracted from the tumor and tumor ring regions. The reproducibility of features from images converted using DNN and interpolation schemes were assessed using the concordance correlation coefficients (CCC) with the cutoff of 0.85. The paired t-test and Mann-Whitney U test were used to compare the evaluation metrics, where appropriate.Results.CT images of 108 patients were used for training (n = 63), validation (n = 11) and testing (n = 34). The DNN method showed significantly higher PSNR and SSIM values (p < 0.05) than interpolation-based methods. The DNN method also showed a significantly higher CCC value than interpolation-based methods. For features in the tumor region, compared with the cubic interpolation approach, the reproducible features increased from 393 (82%) to 422(88%) for the conversion of 3-1 mm, and from 305(64%) to 353(74%) for the conversion of 5-1 mm. For features in the tumor ring region, the improvement was from 395 (82%) to 431 (90%) and from 290 (60%) to 335 (70%), respectively.Conclusions.The DNN based ST up-sampling approach can improve the reproducibility of CT radiomics features in liver tumors, promoting the standardization of CT radiomics studies in liver cancer.

Intra-scan inter-tissue variability can help harmonize radiomics features in CT

OBJECTIVE

We studied the repeatability and the relative intra-scan variability across acquisition protocols in CT using phantom and unenhanced abdominal series.

METHODS

We used 17 CT scans from the Credence Cartridge Radiomics Phantom database and 20 unenhanced multi-site non-pathologic abdominal patient series for which we measured spleen and liver tissues. We performed multiple measurements in extracting 9 radiomics features. We defined a "tandem" as the measurement of a given tissue (or material) by a given radiomics. For each tandem, we assessed the proportion of the variability attributable to repetitions, acquisition protocols, material, or patient. We analyzed the distribution of the intra-scan correlation between pairs of tandems and checked the impact of correlation coefficient greater than 0.90 in comparing paired and unpaired differences.

RESULTS

The repeatability of radiomics features depends on the measured material; 56% of tandems were highly repeatable. Histogram-derived radiomics were generally less repeatable. Nearly 60% of relative radiomics measurements had a correlation coefficient higher than 0.90 allowing paired measurements to improve reliability in detecting the difference between two materials. The analysis of liver and spleen tissues showed that measurement variability was negligible with respect to other variabilities. As for phantom data, we found that gray level zone length matrix (GLZLM)-derived radiomics and gray level co-occurrence matrix (GLCM)-derived radiomics were the most correlating features. For these features, relative intra-scan measurements improved the detection of different materials or tissues.

CONCLUSIONS

We identified radiomics features for which the intra-scan measurements between tissues are linearly correlated. This property represents an opportunity to improve tissue characterization and inter-site harmonization.

KEY POINTS

• The repeatability of radiomics features on CT depends on the measured material or tissue. • Some tandems of radiomics features/tissues are linearly affected by the variability of acquisition protocols on CT. • Relative intra-scan measurements are an opportunity for improving quantitative imaging on CT.

Radiogenomics in brain, breast, and lung cancer: opportunities and challenges

The field of radiogenomics largely focuses on developing imaging surrogates for genomic signatures and integrating imaging, genomic, and molecular data to develop combined personalized biomarkers for characterizing various diseases. Our study aims to highlight the current state-of-the-art and the role of radiogenomics in cancer research, focusing mainly on solid tumors, and is broadly divided into four sections. The first section reviews representative studies that establish the biologic basis of radiomic signatures using gene expression and molecular profiling information. The second section includes studies that aim to non-invasively predict molecular subtypes of tumors using radiomic signatures. The third section reviews studies that evaluate the potential to augment the performance of established prognostic signatures by combining complementary information encoded by radiomic and genomic signatures derived from cancer tumors. The fourth section includes studies that focus on ascertaining the biological significance of radiomic phenotypes. We conclude by discussing current challenges and opportunities in the field, such as the importance of coordination between imaging device manufacturers, regulatory organizations, health care providers, pharmaceutical companies, academic institutions, and physicians for the effective standardization of the results from radiogenomic signatures and for the potential use of these findings to improve precision care for cancer patients.

Systematic Review on the Association of Radiomics with Tumor Biological Endpoints

Radiomics supposes an alternative non-invasive tumor characterization tool, which has experienced increased interest with the advent of more powerful computers and more sophisticated machine learning algorithms. Nonetheless, the incorporation of radiomics in cancer clinical-decision support systems still necessitates a thorough analysis of its relationship with tumor biology. Herein, we present a systematic review focusing on the clinical evidence of radiomics as a surrogate method for tumor molecular profile characterization. An extensive literature review was conducted in PubMed, including papers on radiomics and a selected set of clinically relevant and commonly used tumor molecular markers. We summarized our findings based on different cancer entities, additionally evaluating the effect of different modalities for the prediction of biomarkers at each tumor site. Results suggest the existence of an association between the studied biomarkers and radiomics from different modalities and different tumor sites, even though a larger number of multi-center studies are required to further validate the reported outcomes.

Reproducibility of lung nodule radiomic features: Multivariable and univariable investigations that account for interactions between CT acquisition and reconstruction parameters

PURPOSE

Recent studies have demonstrated a lack of reproducibility of radiomic features in response to variations in CT parameters. In addition, reproducibility of radiomic features has not been well established in clinical datasets. We aimed to investigate the effects of a wide range of CT acquisition and reconstruction parameters on radiomic features in a realistic setting using clinical low dose lung cancer screening cases. We performed univariable and multivariable explorations to consider the effects of individual parameters and the simultaneous interactions between three different acquisition/reconstruction parameters of radiation dose level, reconstructed slice thickness, and kernel.

METHOD

A cohort of 89 lung cancer screening patients were collected that each had a solid lung nodule >4mm diameter. A computational pipeline was used to perform a simulation of dose reduction of the raw projection data, collected from patient scans. This was followed by reconstruction of raw data with weighted filter back projection (wFBP) algorithm and automatic lung nodule detection and segmentation using a computer-aided detection tool. For each patient, 36 different image datasets were created corresponding to dose levels of 100%, 50%, 25%, and 10% of the original dose level, three slice thicknesses of 0.6 mm, 1 mm, and 2 mm, as well as three reconstruction kernels of smooth, medium, and sharp. For each nodule, 226 well-known radiomic features were calculated at each image condition. The reproducibility of radiomic features was first evaluated by measuring the intercondition agreement of the feature values among the 36 image conditions. Then in a series of univariable analyses, the impact of individual CT parameters was assessed by selecting subsets of conditions with one varying and two constant CT parameters. In each subset, intraparameter agreements were assessed. Overall concordance correlation coefficient (OCCC) served as the measure of agreement. An OCCC ≥ 0.9 implied strong agreement and reproducibility of radiomic features in intercondition or intraparameter comparisons. Furthermore, the interaction of CT parameters in impacting radiomic feature values was investigated via ANOVA.

RESULTS

All included radiomic features lacked intercondition reproducibility (OCCC < 0.9) among all the 36 conditions. Out of 226 radiomic features analyzed, only 17 and 18 features were considered reproducible (OCCC ≥ 0.9) to dose and kernel variation, respectively, within the corresponding condition subsets. Slice thickness demonstrated the largest impact on radiomic feature values where only one to five features were reproducible at a few condition subsets. ANOVA revealed significant interactions (P < 0.05) between CT parameters affecting the variability of >50% of radiomic features.

CONCLUSION

We systematically explored the multidimensional space of CT parameters in affecting lung nodule radiomic features. Univariable and multivariable analyses of this study not only showed the lack of reproducibility of the majority of radiomic features but also revealed existing interactions among CT parameters, meaning that the effect of individual CT parameters on radiomic features can be conditional upon other CT acquisition and reconstruction parameters. Our findings advise on careful radiomic feature selection and attention to the inclusion criteria for CT image acquisition protocols within the datasets of radiomic studies.

Glyphosate exposure induces synaptic impairment in hippocampal neurons and cognitive deficits in developing rats

Glyphosate is the active ingredient of several widely used herbicide formulations. Studies based on Glyphosate exposure in different experimental models have suggested that the nervous system represented a key target for its toxicity. Previously, we demonstrated that exposure to glyphosate during gestation induces deficits on behavioral and cognitive function in rats. The aim of the present work was to examine whether cognitive dysfunction induced by Glyphosate was connected to changes on synapse formation and maturation. To understand how glyphosate affects synaptic assembly, we performed in vitro assays on cultured hippocampal neurons that were exposed to the herbicide (0.5 or 1 mg/mL) for 5 or 10 days. Biochemical and immunocytochemical approaches revealed that Glyphosate treated neurons showed a decrease on dendritic complexity and synaptic spine formation and maturation. Moreover, results indicated that Glyphosate decreased synapse formation in hippocampal neurons. To evaluate these effects in vivo, pup rats were treated with 35 or 70 mg/kg of Glyphosate from PND 7 to PND 27, every 48 h. Results indicated that Glyphosate postnatal exposure induced cognitive impairments, since recognition and spatial memory were altered. To go further, we evaluated synaptic protein expression and synaptic organization in hippocampus. Images revealed that Glyphosate treatment downregulates synapsin-1, PSD-95, and CaMKII expression, and also decreased PSD-95 clustering in hippocampus. Taken together, these findings demonstrate for the first time that Glyphosate exposure affects synaptic assembly and reduced synaptic protein expression in hippocampus and that likely triggers the impairment of cognitive function and neuronal connectivity.

Uncontrolled Confounders May Lead to False or Overvalued Radiomics Signature: A Proof of Concept Using Survival Analysis in a Multicenter Cohort of Kidney Cancer

Purpose

We aimed to explore potential confounders of prognostic radiomics signature predicting survival outcomes in clear cell renal cell carcinoma (ccRCC) patients and demonstrate how to control for them.

Materials and Methods

Preoperative contrast enhanced abdominal CT scan of ccRCC patients along with pathological grade/stage, gene mutation status, and survival outcomes were retrieved from The Cancer Imaging Archive (TCIA)/The Cancer Genome Atlas-Kidney Renal Clear Cell Carcinoma (TCGA-KIRC) database, a publicly available dataset. A semi-automatic segmentation method was applied to segment ccRCC tumors, and 1,160 radiomics features were extracted from each segmented tumor on the CT images. Non-parametric principal component decomposition (PCD) and unsupervised hierarchical clustering were applied to build the radiomics signature models. The factors confounding the radiomics signature were investigated and controlled sequentially. Kaplan-Meier curves and Cox regression analyses were performed to test the association between radiomics signatures and survival outcomes.

Results

183 patients of TCGA-KIRC cohort with available imaging, pathological, and clinical outcomes were included in this study. All 1,160 radiomics features were included in the first radiomics signature. Three additional radiomics signatures were then modelled in successive steps removing redundant radiomics features first, removing radiomics features biased by CT slice thickness second, and removing radiomics features dependent on tumor size third. The final radiomics signature model was the most parsimonious, unbiased by CT slice thickness, and independent of tumor size. This final radiomics signature stratified the cohort into radiomics phenotypes that are different by cancer-specific and recurrence-free survival; HR (95% CI) = 3.0 (1.5-5.7), p <0.05 and HR (95% CI) = 6.6 (3.1-14.1), p <0.05, respectively.

Conclusion

Radiomics signature can be confounded by multiple factors, including feature redundancy, image acquisition parameters like slice thickness, and tumor size. Attention to and proper control for these potential confounders are necessary for a reliable and clinically valuable radiomics signature.

The impact of the variation of imaging parameters on the robustness of Computed Tomography radiomic features: A review

The field of radiomics is at the forefront of personalized medicine. However, there is concern that high variation in imaging parameters will impact robustness of radiomic features and subsequently the performance of the predictive models built upon them. Therefore, our review aims to evaluate the impact of imaging parameters on the robustness of radiomic features. We also provide insights into the validity and discrepancy of different methodologies applied to investigate the robustness of radiomic features. We selected 47 papers based on our predefined inclusion criteria and grouped these papers by the imaging parameter under investigation: (i) scanner parameters, (ii) acquisition parameters and (iii) reconstruction parameters. Our review highlighted that most of the imaging parameters are disruptive parameters, and shape along with First order statistics were reported as the most robust radiomic features against variation in imaging parameters. This review identified inconsistencies related to the methodology of the reviewed studies such as the metrics used for robustness, the feature extraction techniques, the reporting style, and their outcome inclusion. We hope this review will aid the scientific community in conducting research in a way that is more reproducible and avoids the pitfalls of previous analyses.

A computed tomography (CT)-derived radiomics approach for predicting primary co-mutations involving TP53 and epidermal growth factor receptor (EGFR) in patients with advanced lung adenocarcinomas (LUAD)

Background

Epidermal growth factor receptor (EGFR) co-mutated with TP53 could reduce responsiveness to tyrosine kinase inhibitors (TKIs) and worsen patients' prognosis compared to TP53 wild type patients in.

EGFR

mutated lung adenocarcinomas (LUAD). To identify this genetically unique subset prior to treatment through computed tomography (CT) images had not been reported yet.

Methods

Stage III and IV LUAD with known mutation status of EGFR and TP53 from The First Affiliated Hospital of Sun Yat-sen University (May 1, 2017 to June 1, 2020) were collected. Characteristics of pretreatment enhanced-CT images were analyzed. One-versus-one was used as the multiclass classification strategy to distinguish the three subtypes of co-mutations: EGFR ⁺ & TP53 ⁺, EGFR ⁺ & TP53 ^-, EGFR ^-. The clinical model, semantic model, radiomics model and integrated model were built. Area under the receiver-operating characteristic curves (AUCs) were used to evaluate the prediction efficacy.

Results

A total of 199 patients were enrolled, including 83 (42%) cases of EGFR ^-, 55 (28%) cases of EGFR ⁺ & TP53 ⁺, 61 (31%) cases of EGFR ⁺ & TP53 ^-. Among the four different models, the integrated model displayed the best performance for all the three subtypes of co-mutations: EGFR ^- (AUC, 0.857; accuracy, 0.817; sensitivity, 0.998; specificity, 0.663), EGFR ⁺ & TP53 ⁺ (AUC, 0.791; accuracy, 0.758; sensitivity, 0.762; specificity, 0.783), EGFR ⁺ & TP53 ^- (AUC, 0.761; accuracy, 0.813; sensitivity, 0.594; specificity, 0.977). The radiomics model was slightly inferior to the integrated model. The results for the clinical and the semantic models were dissatisfactory, with AUCs less than 0.700 for all the three subtypes.

Conclusions

CT imaging based artificial intelligence (AI) is expected to distinguish co-mutation status involving TP53 and EGFR. The proposed integrated model may serve as an important alternative marker for preselecting patients who will be adaptable to and sensitive to TKIs.

Understanding Sources of Variation to Improve the Reproducibility of Radiomics

Radiomics is the method of choice for investigating the association between cancer imaging phenotype, cancer genotype and clinical outcome prediction in the era of precision medicine. The fast dispersal of this new methodology has benefited from the existing advances of the core technologies involved in radiomics workflow: image acquisition, tumor segmentation, feature extraction and machine learning. However, despite the rapidly increasing body of publications, there is no real clinical use of a developed radiomics signature so far. Reasons are multifaceted. One of the major challenges is the lack of reproducibility and generalizability of the reported radiomics signatures (features and models). Sources of variation exist in each step of the workflow; some are controllable or can be controlled to certain degrees, while others are uncontrollable or even unknown. Insufficient transparency in reporting radiomics studies further prevents translation of the developed radiomics signatures from the bench to the bedside. This review article first addresses sources of variation, which is illustrated using demonstrative examples. Then, it reviews a number of published studies and progresses made to date in the investigation and improvement of feature reproducibility and model performance. Lastly, it discusses potential strategies and practical considerations to reduce feature variability and improve the quality of radiomics study. This review focuses on CT image acquisition, tumor segmentation, quantitative feature extraction, and the disease of lung cancer.

The combined use of 2D scout and 3D axial CT images to accurately determine the catheter tips for high-dose-rate brachytherapy plans

Purpose

To develop a method combining CT scout images with axial images to improve the localization accuracy of catheter tips in high‐dose‐rate (HDR) brachytherapy treatments.

Materials and Methods

CT scout images were utilized along with conventionally reconstructed axial images to aid the localization of catheter tips used during HDR treatment planning. A method was developed to take advantage of the finer image resolution of the scout images to more precisely identify the tip coordinates. The accuracies of this method were compared with the conventional method based on the axial CT images alone, for various slice thicknesses, in a computed tomography dose index (CTDI) head phantom. A clinical case which involved multiple interstitial catheters was also selected for the evaluation of this method. Locations of the catheter tips were reconstructed with the conventional CT‐based method and this newly developed method, respectively. Location coordinates obtained via both methods were quantitatively compared.

Results

Combination of the scout and axial CT images improved the accuracy of identification and reconstruction of catheter tips along the longitudinal direction (i.e., head‐to‐foot direction, more or less parallel to the catheter tracks), compared to relying on the axial CT images alone. The degree of improvement was dependent on CT slice thickness. For the clinical patient case, the coordinate differences of the reconstructed catheter tips were 2.6 mm ± 0.9 mm in the head‐to‐foot direction, 0.4 mm ± 0.2 mm in the left‐to‐right direction, and 0.6 mm ± 0.2 mm in the anterior‐to‐posterior direction, respectively.

Conclusion

Combining CT scout and axial images demonstrates the ability to provide a more accurate identification and reconstruction of the interstitial catheter tips for HDR brachytherapy treatment, especially in the longitudinal direction. The method developed in this work has the potential to be implemented clinically together with automatic segmentation in modern brachytherapy treatment planning systems, in order to improve the reconstruction accuracy of HDR catheters.

Robustness of magnetic resonance radiomic features to pixel size resampling and interpolation in patients with cervical cancer

Background

Radiomics is a promising field in oncology imaging. However, the implementation of radiomics clinically has been limited because its robustness remains unclear. Previous CT and PET studies suggested that radiomic features were sensitive to variations in pixel size and slice thickness of the images. The purpose of this study was to assess robustness of magnetic resonance (MR) radiomic features to pixel size resampling and interpolation in patients with cervical cancer.

Methods

This retrospective study included 254 patients with a pathological diagnosis of cervical cancer stages IB to IVA who received definitive chemoradiation at our institution between January 2006 and June 2020. Pretreatment MR scans were analyzed. Each region of cervical cancer was segmented on the axial gadolinium-enhanced T1- and T2-weighted images; 107 radiomic features were extracted. MR scans were interpolated and resampled using various slice thicknesses and pixel spaces. Intraclass correlation coefficients (ICCs) were calculated between the original images and images that underwent pixel size resampling (OP), interpolation (OI), or pixel size resampling and interpolation (OP+I) as well as among processed image sets with various pixel spaces (P), various slice thicknesses (I), and both (P + I).

Results

After feature standardization, ≥86.0% of features showed good robustness when compared between the original and processed images (OP, OI, and OP+I) and ≥ 88.8% of features showed good robustness when processed images were compared (P, I, and P + I). Although most first-order, shape, and texture features showed good robustness, GLSZM small-area emphasis-related features and NGTDM strength were sensitive to variations in pixel size and slice thickness.

Conclusion

Most MR radiomic features in patients with cervical cancer were robust after pixel size resampling and interpolation following the feature standardization process. The understanding regarding the robustness of individual features after pixel size resampling and interpolation could help future radiomics research.

Supplementary Information

The online version contains supplementary material available at 10.1186/s40644-021-00388-5.

Robust radiogenomics approach to the identification of EGFR mutations among patients with NSCLC from three different countries using topologically invariant Betti numbers

OBJECTIVES

To propose a novel robust radiogenomics approach to the identification of epidermal growth factor receptor (EGFR) mutations among patients with non-small cell lung cancer (NSCLC) using Betti numbers (BNs).

MATERIALS AND METHODS

Contrast enhanced computed tomography (CT) images of 194 multi-racial NSCLC patients (79 EGFR mutants and 115 wildtypes) were collected from three different countries using 5 manufacturers' scanners with a variety of scanning parameters. Ninety-nine cases obtained from the University of Malaya Medical Centre (UMMC) in Malaysia were used for training and validation procedures. Forty-one cases collected from the Kyushu University Hospital (KUH) in Japan and fifty-four cases obtained from The Cancer Imaging Archive (TCIA) in America were used for a test procedure. Radiomic features were obtained from BN maps, which represent topologically invariant heterogeneous characteristics of lung cancer on CT images, by applying histogram- and texture-based feature computations. A BN-based signature was determined using support vector machine (SVM) models with the best combination of features that maximized a robustness index (RI) which defined a higher total area under receiver operating characteristics curves (AUCs) and lower difference of AUCs between the training and the validation. The SVM model was built using the signature and optimized in a five-fold cross validation. The BN-based model was compared to conventional original image (OI)- and wavelet-decomposition (WD)-based models with respect to the RI between the validation and the test.

RESULTS

The BN-based model showed a higher RI of 1.51 compared with the models based on the OI (RI: 1.33) and the WD (RI: 1.29).

CONCLUSION

The proposed model showed higher robustness than the conventional models in the identification of EGFR mutations among NSCLC patients. The results suggested the robustness of the BN-based approach against variations in image scanner/scanning parameters.

Radiomic features based on Hessian index for prediction of prognosis in head-and-neck cancer patients

This study demonstrated the usefulness of radiomic features based on the Hessian index of differential topology for the prediction of prognosis prior to treatment in head-and-neck (HN) cancer patients. The Hessian index, which can indicate tumor heterogeneity with convex, concave, and other points (saddle points), was calculated as the number of negative eigenvalues of the Hessian matrix at each voxel on computed tomography (CT) images. Three types of signatures were constructed in a training cohort (n = 126), one type each from CT conventional features, Hessian index features, and combined features from the conventional and index feature sets. The prognostic value of the signatures were evaluated using statistically significant difference (p value, log-rank test) to compare the survival curves of low- and high-risk groups. In a test cohort (n = 68), the p values of the models built with conventional, index, combined features, and clinical variables were 2.95 [Formula: see text] 10-2, 1.85 [Formula: see text] 10-2, 3.17 [Formula: see text] 10-2, and 1.87 [Formula: see text] 10-3, respectively. When the features were integrated with clinical variables, the p values of conventional, index, and combined features were 3.53 [Formula: see text] 10-3, 1.28 [Formula: see text] 10-3, and 1.45 [Formula: see text] 10-3, respectively. This result indicates that index features could provide more prognostic information than conventional features and further increase the prognostic value of clinical variables in HN cancer patients.

The Usefulness of Pretreatment MR-Based Radiomics on Early Response of Neoadjuvant Chemotherapy in Patients With Locally Advanced Nasopharyngeal Carcinoma

The aim of this study was to explore the predictive role of pretreatment MRI-based radiomics on early response of neoadjuvant chemotherapy (NAC) in locoregionally advanced nasopharyngeal carcinoma (NPC) patients. Between January 2016 and December 2016, a total of 108 newly diagnosed NPC patients who were hospitalized in the Cancer Hospital of the University of Chinese Academy of Sciences were reviewed. All patients had complete data of enhanced MR of nasopharynx before treatment, and then received two to three cycles of TP-based NAC. After 2 cycles of NAC, enhanced MR of nasopharynx was conducted again. Compared with the enhanced MR images before treatment, the response after NAC was evaluated. According to the evaluation criteria of RECIST1.1, 108 cases were divided into two groups: 52 cases for the NAC-sensitive group and 56 cases for the NAC-resistance group. ITK-SNAP software was used to manually sketch and segment the region of interest (ROI) of nasopharyngeal tumor on the MR enhanced T1WI sequence image. The parameters were analyzed and extracted by using AI Kit software. ANOVA/MW test, correlation analysis, and LASSO were used to select texture features. We used multivariate logistic regressions to select texture features and establish a predictive model. The ROC curve was used to evaluate the efficiency of the predictive model. A total of 396 texture features were obtained by using feature calculation. After all features were screened, we selected two features including ClusterShade_angle135_offset4 and Correlation_AllDirection_offshe1_SD. Based on these two features, we established a predictive model by using multivariate logistic regression. The AUC of the two features used alone (0.804, 95% CI=0.6020.932; 0.762, 95% CI=0.5560.905) was smaller than the combination of these two features (0.905, 95% CI=0.7240.984, p=0.0005). Moreover, the sensitivity values of the two features used alone and the combined use were 92.9%, 51.7%, and 85.7%, respectively, while the specificity values were 66.7%, 91.7%, and 83.3%, respectively, in the early response of NAC for NPC. The predictive model based on MRI-enhanced sequence imaging could distinguish the sensitivity and resistance to NAC and provide new biomarkers for the early prediction of the curative effect in NPC patients.

CT Radiomics in Predicting EGFR Mutation in Non-small Cell Lung Cancer: A Single Institutional Study

Objective: To evaluate the value of CT radiomics in predicting the epidermal growth factor receptor (EGFR) mutation of patients with non-small cell lung cancer (NSCLC), and combing with the clinical characteristic to construct the prediction model.

Methods: Sixty-seven cases of NSCLC confirmed by pathology were enrolled. The pre-treatment chest CT enhanced images were used in Radiomics analysis. Two experienced radiologists delineated the region of interest (ROI) on open source software 3D-Slicer. The feature of ROI was extracted by Pyradiomics software package and a total of 849 features were extracted. By calculating Pearson correlation coefficient between pair-wise features and LASSO method for feature screening. The prediction model was constructed by logical regression, diagnostic efficacy of the model by the area under the receiver operating characteristic (ROC) curve was calculated.

Results: Based on clinical model and the radiomics model, the AUC under the ROC was 0.8387 and 0.8815, respectively. The model combining clinical and radiomics features perfect best, the AUC under the ROC was 0.9724, the sensitivity and specificity were 85.3 and 90.9%, respectively.

Conclusions: Compared with clinical features or radiomics features alone, the model constructed by combining clinical and pre-treatment chest enhanced CT features may show more utility for improved patient stratification in EGFR mutation and EGFR wild.

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.

The application of artificial intelligence and radiomics in lung cancer

Lung cancer is one of the most leading causes of death throughout the world, and there is an urgent requirement for the precision medical management of it. Artificial intelligence (AI) consisting of numerous advanced techniques has been widely applied in the field of medical care. Meanwhile, radiomics based on traditional machine learning also does a great job in mining information through medical images. With the integration of AI and radiomics, great progress has been made in the early diagnosis, specific characterization, and prognosis of lung cancer, which has aroused attention all over the world. In this study, we give a brief review of the current application of AI and radiomics for precision medical management in lung cancer.

Screening of optimal reference genes for qRT-PCR and preliminary exploration of cold resistance mechanisms in Prunus mume and Prunus sibirica varieties

Prunus sibirica and Prunus mume are closely related plant species that differ in cold tolerance. Hybrids of P. sibirica and true mume, belonging to the apricot mei group, inherited strong cold resistance from P. sibirica. These materials are favourable for research on the molecular mechanisms of cold resistance. However, no suitable reference genes have been identified for analysing gene expression patterns between P. sibirica and P. mume. Ten candidate reference genes were assessed, namely, actins (ACT2-1, ACT2-2, ACT2-3, ACT2-4), protein phosphatase 2A-1 (PP2A-1), ubiquitins (UBQ2, UBQ3), ubiquitin extension protein (UBQ1) and tubulins (TUB1, TUB2), with four distinct algorithms (geNorm, NormFinder, BestKeeper and RefFinder). UBQ2 was recognized as the best reference gene in stems and buds across materials (P. sibirica; 'Xiaohong Zhusha', 'Beijing Yudie', and 'Xiao Lve' for true mume; and 'Dan Fenghou', 'Fenghou', and 'Yanxing' for apricot mei) under cold stress. In addition, the temporal and spatial expression patterns of PmCBF6 and PmLEA10 among seven varieties during winter periods were analysed using UBQ2 as a reference gene. The expression differed significantly among cultivars, which may contribute to their differences in cold tolerance. This paper confirmed the strong cold tolerance of apricot mei. And the best internal reference gene suitable for seven varieties was selected: UBQ2. Based on the above results, the expression of PmCBF6 and PmLEA10 genes during wintering in seven varieties was analysed. The molecular mechanisms of cold resistance were found to be possibly different in different varieties of P. sibirica and P. mume.

Reinventing radiation therapy with machine learning and imaging bio-markers (radiomics): State-of-the-art, challenges and perspectives

Radiation therapy is a pivotal cancer treatment that has significantly progressed over the last decade due to numerous technological breakthroughs. Imaging is now playing a critical role on deployment of the clinical workflow, both for treatment planning and treatment delivery. Machine-learning analysis of predefined features extracted from medical images, i.e. radiomics, has emerged as a promising clinical tool for a wide range of clinical problems addressing drug development, clinical diagnosis, treatment selection and implementation as well as prognosis. Radiomics denotes a paradigm shift redefining medical images as a quantitative asset for data-driven precision medicine. The adoption of machine-learning in a clinical setting and in particular of radiomics features requires the selection of robust, representative and clinically interpretable biomarkers that are properly evaluated on a representative clinical data set. To be clinically relevant, radiomics must not only improve patients' management with great accuracy but also be reproducible and generalizable. Hence, this review explores the existing literature and exposes its potential technical caveats, such as the lack of quality control, standardization, sufficient sample size, type of data collection, and external validation. Based upon the analysis of 165 original research studies based on PET, CT-scan, and MRI, this review provides an overview of new concepts, and hypotheses generating findings that should be validated. In particular, it describes evolving research trends to enhance several clinical tasks such as prognostication, treatment planning, response assessment, prediction of recurrence/relapse, and prediction of toxicity. Perspectives regarding the implementation of an AI-based radiotherapy workflow are presented.