Measuring Computed Tomography Scanner Variability of Radiomics Features. Invest Radiol. 2015;50:757-765. [PMID: 26115366 DOI: 10.1097/rli.0000000000000180]

Deep Learning Auto-Segmentation Network for Pediatric Computed Tomography Data Sets: Can We Extrapolate From Adults?

PURPOSE

Artificial intelligence (AI)-based auto-segmentation models hold promise for enhanced efficiency and consistency in organ contouring for adaptive radiation therapy and radiation therapy planning. However, their performance on pediatric computed tomography (CT) data and cross-scanner compatibility remain unclear. This study aimed to evaluate the performance of AI-based auto-segmentation models trained on adult CT data when applied to pediatric data sets and explore the improvement in performance gained by including pediatric training data. It also examined their ability to accurately segment CT data acquired from different scanners.

METHODS AND MATERIALS

Using the nnU-Net framework, segmentation models were trained on data sets of adult, pediatric, and combined CT scans for 7 pelvic/thoracic organs. Each model was trained on 290 to 300 cases per category and organ. Training data sets included a combination of clinical data and several open repositories. The study incorporated a database of 459 pediatric (0-16 years) CT scans and 950 adults (>18 years), ensuring all scans had human expert ground-truth contours of the selected organs. Performance was evaluated based on Dice similarity coefficients (DSC) of the model-generated contours.

RESULTS

AI models trained exclusively on adult data underperformed on pediatric data, especially for the 0 to 2 age group: mean DSC was below 0.5 for the bladder and spleen. The addition of pediatric training data demonstrated significant improvement for all age groups, achieving a mean DSC of above 0.85 for all organs in every age group. Larger organs like the liver and kidneys maintained consistent performance for all models across age groups. No significant difference emerged in the cross-scanner performance evaluation, suggesting robust cross-scanner generalization.

CONCLUSIONS

For optimal segmentation across age groups, it is important to include pediatric data in the training of segmentation models. The successful cross-scanner generalization also supports the real-world clinical applicability of these AI models. This study emphasizes the significance of data set diversity in training robust AI systems for medical image interpretation tasks.

Machine learning prediction model for gray-level co-occurrence matrix features of synchronous liver metastasis in colorectal cancer

BACKGROUND

Synchronous liver metastasis (SLM) is a significant contributor to morbidity in colorectal cancer (CRC). There are no effective predictive device integration algorithms to predict adverse SLM events during the diagnosis of CRC.

AIM

To explore the risk factors for SLM in CRC and construct a visual prediction model based on gray-level co-occurrence matrix (GLCM) features collected from magnetic resonance imaging (MRI).

METHODS

Our study retrospectively enrolled 392 patients with CRC from Yichang Central People’s Hospital from January 2015 to May 2023. Patients were randomly divided into a training and validation group (3:7). The clinical parameters and GLCM features extracted from MRI were included as candidate variables. The prediction model was constructed using a generalized linear regression model, random forest model (RFM), and artificial neural network model. Receiver operating characteristic curves and decision curves were used to evaluate the prediction model.

RESULTS

Among the 392 patients, 48 had SLM (12.24%). We obtained fourteen GLCM imaging data for variable screening of SLM prediction models. Inverse difference, mean sum, sum entropy, sum variance, sum of squares, energy, and difference variance were listed as candidate variables, and the prediction efficiency (area under the curve) of the subsequent RFM in the training set and internal validation set was 0.917 [95% confidence interval (95%CI): 0.866-0.968] and 0.09 (95%CI: 0.858-0.960), respectively.

CONCLUSION

A predictive model combining GLCM image features with machine learning can predict SLM in CRC. This model can assist clinicians in making timely and personalized clinical decisions.

The current landscape of machine learning-based radiomics in arteriovenous malformations: a systematic review and radiomics quality score assessment

Background

Arteriovenous malformations (AVMs) are rare vascular anomalies involving a disorganization of arteries and veins with no intervening capillaries. In the past 10 years, radiomics and machine learning (ML) models became increasingly popular for analyzing diagnostic medical images. The goal of this review was to provide a comprehensive summary of current radiomic models being employed for the diagnostic, therapeutic, prognostic, and predictive outcomes in AVM management.

Methods

A systematic literature review was conducted according to the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) 2020 guidelines, in which the PubMed and Embase databases were searched using the following terms: (cerebral OR brain OR intracranial OR central nervous system OR spine OR spinal) AND (AVM OR arteriovenous malformation OR arteriovenous malformations) AND (radiomics OR radiogenomics OR machine learning OR artificial intelligence OR deep learning OR computer-aided detection OR computer-aided prediction OR computer-aided treatment decision). A radiomics quality score (RQS) was calculated for all included studies.

Results

Thirteen studies were included, which were all retrospective in nature. Three studies (23%) dealt with AVM diagnosis and grading, 1 study (8%) gauged treatment response, 8 (62%) predicted outcomes, and the last one (8%) addressed prognosis. No radiomics model had undergone external validation. The mean RQS was 15.92 (range: 10-18).

Conclusion

We demonstrated that radiomics is currently being studied in different facets of AVM management. While not ready for clinical use, radiomics is a rapidly emerging field expected to play a significant future role in medical imaging. More prospective studies are warranted to determine the role of radiomics in the diagnosis, prediction of comorbidities, and treatment selection in AVM management.

Towards equitable AI in oncology

Artificial intelligence (AI) stands at the threshold of revolutionizing clinical oncology, with considerable potential to improve early cancer detection and risk assessment, and to enable more accurate personalized treatment recommendations. However, a notable imbalance exists in the distribution of the benefits of AI, which disproportionately favour those living in specific geographical locations and in specific populations. In this Perspective, we discuss the need to foster the development of equitable AI tools that are both accurate in and accessible to a diverse range of patient populations, including those in low-income to middle-income countries. We also discuss some of the challenges and potential solutions in attaining equitable AI, including addressing the historically limited representation of diverse populations in existing clinical datasets and the use of inadequate clinical validation methods. Additionally, we focus on extant sources of inequity including the type of model approach (such as deep learning, and feature engineering-based methods), the implications of dataset curation strategies, the need for rigorous validation across a variety of populations and settings, and the risk of introducing contextual bias that comes with developing tools predominantly in high-income countries.

The impact of the combat method on radiomics feature compensation and analysis of scanners from different manufacturers

BACKGROUND

This study investigated whether the Combat compensation method can remove the variability of radiomic features extracted from different scanners, while also examining its impact on the subsequent predictive performance of machine learning models.

MATERIALS AND METHODS

135 CT images of Credence Cartridge Radiomic phantoms were collected and screened from three scanners manufactured by Siemens, Philips, and GE. 100 radiomic features were extracted and 20 radiomic features were screened according to the Lasso regression method. The radiomic features extracted from the rubber and resin-filled regions in the cartridges were labeled into different categories for evaluating the performance of the machine learning model. Radiomics features were divided into three groups based on the different scanner manufacturers. The radiomic features were randomly divided into training and test sets with a ratio of 8:2. Five machine learning models (lasso, logistic regression, random forest, support vector machine, neural network) were employed to evaluate the impact of Combat on radiomic features. The variability among radiomic features were assessed using analysis of variance (ANOVA) and principal component analysis (PCA). Accuracy, precision, recall, and area under the receiver curve (AUC) were used as evaluation metrics for model classification.

RESULTS

The principal component and ANOVA analysis results show that the variability of different scanner manufacturers in radiomic features was removed (P˃0.05). After harmonization with the Combat algorithm, the distributions of radiomic features were aligned in terms of location and scale. The performance of machine learning models for classification improved, with the Random Forest model showing the most significant enhancement. The AUC value increased from 0.88 to 0.92.

CONCLUSIONS

The Combat algorithm has reduced variability in radiomic features from different scanners. In the phantom CT dataset, it appears that the machine learning model's classification performance may have improved after Combat harmonization. However, further investigation and validation are required to fully comprehend Combat's impact on radiomic features in medical imaging.

Comparing quantitative image parameters between animal and clinical CT-scanners: a translational phantom study analysis

Purpose

This study compares phantom-based variability of extracted radiomics features from scans on a photon counting CT (PCCT) and an experimental animal PET/CT-scanner (Albira II) to investigate the potential of radiomics for translation from animal models to human scans. While oncological basic research in animal PET/CT has allowed an intrinsic comparison between PET and CT, but no 1:1 translation to a human CT scanner due to resolution and noise limitations, Radiomics as a statistical and thus scale-independent method can potentially close the critical gap.

Methods

Two phantoms were scanned on a PCCT and animal PET/CT-scanner with different scan parameters and then the radiomics parameters were extracted. A Principal Component Analysis (PCA) was conducted. To overcome the limitation of a small dataset, a data augmentation technique was applied. A Ridge Classifier was trained and a Feature Importance- and Cluster analysis was performed.

Results

PCA and Cluster Analysis shows a clear differentiation between phantom types while emphasizing the comparability of both scanners. The Ridge Classifier exhibited a strong training performance with 93% accuracy, but faced challenges in generalization with a test accuracy of 62%.

Conclusion

These results show that radiomics has great potential as a translational tool between animal models and human routine diagnostics, especially using the novel photon counting technique. This is another crucial step towards integration of radiomics analysis into clinical practice.

A new computed tomography score-based staging for melioidosis pneumonia to predict progression

Background

Melioidosis pneumonia, caused by the bacterium Burkholderia pseudomallei, is a serious infectious disease prevalent in tropical regions. Chest computed tomography (CT) has emerged as a valuable tool for assessing the severity and progression of lung involvement in melioidosis pneumonia. However, there persists a need for the quantitative assessment of CT characteristics and staging methodologies to precisely anticipate disease progression. This study aimed to quantitatively extract CT features and evaluate a CT score-based staging system in predicting the progression of melioidosis pneumonia.

Methods

This study included 97 patients with culture-confirmed melioidosis pneumonia who presented between January 2002 and December 2021. Lung segmentation and annotation of lesions (consolidation, nodules, and cavity) were used for feature extraction. The features, including the involved area, amount, and intensity, were extracted. The CT scores of the lesion features were defined by the feature importance weight and qualitative stage of melioidosis pneumonia. Gaussian process regression (GPR) was used to predict patients with severe or critical melioidosis pneumonia according to CT scores.

Results

The melioidosis pneumonia stages included acute stage (0-7 days), subacute stage (8-28 days), and chronic stage (>28 days). In the acute stage, the CT scores of all patients ranged from 2.5 to 6.5. In the subacute stage, the CT scores for the severe and mild patients were 3.0-7.0 and 2.0-5.0, respectively. In the chronic stage, the CT score of the mild patients fluctuated approximately between 2.5 and 3.5 in a linear distribution. Consolidation was the most common type of lung lesion in those with melioidosis pneumonia. Between stages I and II, the percentage of severe scans with nodules dropped from 72.22% to 47.62% (P<0.05), and the percentage of severe scans with cavities significantly increased from 16.67% to 57.14% (P<0.05). The GPR optimization function yielded area under the receiver operating characteristic curves of 0.71 for stage I, 0.92 for stage II, and 0.87 for all stages.

Conclusions

In patients with melioidosis pneumonia, it is reasonable to divide the period (the whole progression of melioidosis pneumonia) into three stages to determine the prognosis.

The textures of sarcoidosis: quantifying lung disease through variograms

Objective

Sarcoidosis is a granulomatous disease affecting the lungs in over 90% of patients. Qualitative assessment of chest CT by radiologists is standard clinical practice and reliable quantification of disease from CT would support ongoing efforts to identify sarcoidosis phenotypes. Standard imaging feature engineering techniques such as radiomics suffer from extreme sensitivity to image acquisition and processing, potentially impeding generalizability of research to clinical populations. In this work, we instead investigate approaches to engineering variogram-based features with the intent to identify a robust, generalizable pipeline for image quantification in the study of sarcoidosis.

Approach

For a cohort of more than 300 individuals with sarcoidosis, we investigated 24 feature engineering pipelines differing by decisions for image registration to a template lung, empirical and model variogram estimation methods, and feature harmonization for CT scanner model, and subsequently 48 sets of phenotypes produced through unsupervised clustering. We then assessed sensitivity of engineered features, phenotypes produced through unsupervised clustering, and sarcoidosis disease signal strength to pipeline.

Main results

We found that variogram features had low to mild association with scanner model and associations were reduced by image registration. For each feature type, features were also typically robust to all pipeline decisions except image registration. Strength of disease signal as measured by association with pulmonary function testing and some radiologist visual assessments was strong (optimistic AUC ≈ 0.9, p ≪ 0.0001 in models for architectural distortion, conglomerate mass, fibrotic abnormality, and traction bronchiectasis) and fairly consistent across engineering approaches regardless of registration and harmonization for CT scanner.

Significance

Variogram-based features appear to be a suitable approach to image quantification in support of generalizable research in pulmonary sarcoidosis.

Radiomics in Diagnosis, Grading, and Treatment Response Assessment of Soft Tissue Sarcomas: A Systematic Review and Meta-analysis

RATIONALE AND OBJECTIVES

To evaluate radiomics in soft tissue sarcomas (STSs) for diagnostic accuracy, grading, and treatment response assessment, with a focus on clinical relevance.

METHODS

In this diagnostic accuracy study, radiomics was applied using multiple MRI sequences and AI classifiers, with histopathological diagnosis as the reference standard. Statistical analysis involved meta-analysis, random-effects model, and Deeks' funnel plot asymmetry test.

RESULTS

Among 579 unique titles and abstracts, 24 articles were included in the systematic review, with 21 used for meta-analysis. Radiomics demonstrated a pooled sensitivity of 84% (95% CI: 80-87) and specificity of 63% (95% CI: 56-70), AUC of 0.93 for diagnosis, sensitivity of 84% (95% CI: 82-87) and specificity of 73% (95% CI: 68-77), AUC of 0.91 for grading, and sensitivity of 83% (95% CI: 67-94) and specificity of 67% (95% CI: 59-74), AUC of 0.87 for treatment response assessment.

CONCLUSION

Radiomics exhibits potential for accurate diagnosis, grading, and treatment response assessment in STSs, emphasizing the need for standardization and prospective trials.

CLINICAL RELEVANCE STATEMENT

Radiomics offers precise tools for STS diagnosis, grading, and treatment response assessment, with implications for optimizing patient care and treatment strategies in this complex malignancy.

Influence of image preprocessing on the segmentation-based reproducibility of radiomic features: in vivo experiments on discretization and resampling parameters

PURPOSE

To systematically investigate the impact of image preprocessing parameters on the segmentation-based reproducibility of magnetic resonance imaging (MRI) radiomic features.

METHODS

The MRI scans of 50 patients were included from the multi-institutional Brain Tumor Segmentation 2021 public glioma dataset. Whole tumor volumes were manually segmented by two independent readers, with the participation of eight readers. Radiomic features were extracted from two sequences: T2-weighted (T2) and contrast-enhanced T1-weighted (T1ce). Two methods were considered for discretization: bin count (i.e., relative discretization) and bin width (i.e., absolute discretization). Ten discretization (five for each method) and five resampling parameters were varied while other parameters were fixed. The intraclass correlation coefficient (ICC) was used for reliability analysis based on two commonly used cut-off values (0.75 and 0.90).

RESULTS

Image preprocessing parameters had a significant impact on the segmentation-based reproducibility of radiomic features. The bin width method yielded more reproducible features than the bin count method. In discretization experiments using the bin width on both sequences, according to the ICC cut-off values of 0.75 and 0.90, the rate of reproducible features ranged from 70% to 84% and from 35% to 57%, respectively, with an increasing percentage trend as parameter values decreased (from 84 to 5 for T2; 100 to 6 for T1ce). In the resampling experiments, these ranged from 53% to 74% and from 10% to 20%, respectively, with an increasing percentage trend from lower to higher parameter values (physical voxel size; from 1 x 1 x 1 to 2 x 2 x 2 mm3).

CONCLUSION

The segmentation-based reproducibility of radiomic features appears to be substantially influenced by discretization and resampling parameters. Our findings indicate that the bin width method should be used for discretization and lower bin width and higher resampling values should be used to allow more reproducible features.

Reproducible Radiomics Features from Multi-MRI-Scanner Test-Retest-Study: Influence on Performance and Generalizability of Models

BACKGROUND

Radiomics models trained on data from one center typically show a decline of performance when applied to data from external centers, hindering their introduction into large-scale clinical practice. Current expert recommendations suggest to use only reproducible radiomics features isolated by multiscanner test-retest experiments, which might help to overcome the problem of limited generalizability to external data.

PURPOSE

To evaluate the influence of using only a subset of robust radiomics features, defined in a prior in vivo multi-MRI-scanner test-retest-study, on the performance and generalizability of radiomics models.

STUDY TYPE

Retrospective.

POPULATION

Patients with monoclonal plasma cell disorders. Training set (117 MRIs from center 1); internal test set (42 MRIs from center 1); external test set (143 MRIs from center 2-8).

FIELD STRENGTH/SEQUENCE

1.5T and 3.0T; T1-weighted turbo spin echo.

ASSESSMENT

The task for the radiomics models was to predict plasma cell infiltration, determined by bone marrow biopsy, noninvasively from MRI. Radiomics machine learning models, including linear regressor, support vector regressor (SVR), and random forest regressor (RFR), were trained on data from center 1, using either all radiomics features, or using only reproducible radiomics features. Models were tested on an internal (center 1) and a multicentric external data set (center 2-8).

STATISTICAL TESTS

Pearson correlation coefficient r and mean absolute error (MAE) between predicted and actual plasma cell infiltration. Fisher's z-transformation, Wilcoxon signed-rank test, Wilcoxon rank-sum test; significance level P < 0.05.

RESULTS

When using only reproducible features compared with all features, the performance of the SVR on the external test set significantly improved (r = 0.43 vs. r = 0.18 and MAE = 22.6 vs. MAE = 28.2). For the RFR, the performance on the external test set deteriorated when using only reproducible instead of all radiomics features (r = 0.33 vs. r = 0.44, P = 0.29 and MAE = 21.9 vs. MAE = 20.5, P = 0.10).

CONCLUSION

Using only reproducible radiomics features improves the external performance of some, but not all machine learning models, and did not automatically lead to an improvement of the external performance of the overall best radiomics model.

LEVEL OF EVIDENCE: 3

TECHNICAL EFFICACY

Stage 2.

Identification of CT radiomic features robust to acquisition and segmentation variations for improved prediction of radiotherapy-treated lung cancer patient recurrence

The primary objective of the present study was to identify a subset of radiomic features extracted from primary tumor imaged by computed tomography of early-stage non-small cell lung cancer patients, which remain unaffected by variations in segmentation quality and in computed tomography image acquisition protocol. The robustness of these features to segmentation variations was assessed by analyzing the correlation of feature values extracted from lesion volumes delineated by two annotators. The robustness to variations in acquisition protocol was evaluated by examining the correlation of features extracted from high-dose and low-dose computed tomography scans, both of which were acquired for each patient as part of the stereotactic body radiotherapy planning process. Among 106 radiomic features considered, 21 were identified as robust. An analysis including univariate and multivariate assessments was subsequently conducted to estimate the predictive performance of these robust features on the outcome of early-stage non-small cell lung cancer patients treated with stereotactic body radiation therapy. The univariate predictive analysis revealed that robust features demonstrated superior predictive potential compared to non-robust features. The multivariate analysis indicated that linear regression models built with robust features displayed greater generalization capabilities by outperforming other models in predicting the outcomes of an external validation dataset.

Dose-Incorporated Deep Ensemble Learning for Improving Brain Metastasis Stereotactic Radiosurgery Outcome Prediction

PURPOSE

To develop a novel deep ensemble learning model for accurate prediction of brain metastasis (BM) local control outcomes after stereotactic radiosurgery (SRS).

METHODS AND MATERIALS

A total of 114 brain metastases (BMs) from 82 patients were evaluated, including 26 BMs that developed biopsy-confirmed local failure post-SRS. The SRS spatial dose distribution (Dmap) of each BM was registered to the planning contrast-enhanced T1 (T1-CE) magnetic resonance imaging (MRI). Axial slices of the Dmap, T1-CE, and planning target volume (PTV) segmentation (PTVseg) intersecting the BM center were extracted within a fixed field of view determined by the 60% isodose volume in Dmap. A spherical projection was implemented to transform planar image content onto a spherical surface using multiple projection centers, and the resultant T1-CE/Dmap/PTVseg projections were stacked as a 3-channel variable. Four Visual Geometry Group (VGG-19) deep encoders were used in an ensemble design, with each submodel using a different spherical projection formula as input for BM outcome prediction. In each submodel, clinical features after positional encoding were fused with VGG-19 deep features to generate logit results. The ensemble's outcome was synthesized from the 4 submodel results via logistic regression. In total, 10 model versions with random validation sample assignments were trained to study model robustness. Performance was compared with (1) a single VGG-19 encoder, (2) an ensemble with a T1-CE MRI as the sole image input after projections, and (3) an ensemble with the same image input design without clinical feature inclusion.

RESULTS

The ensemble model achieved an excellent area under the receiver operating characteristic curve (AUCROC: 0.89 ± 0.02) with high sensitivity (0.82 ± 0.05), specificity (0.84 ± 0.11), and accuracy (0.84 ± 0.08) results. This outperformed the MRI-only VGG-19 encoder (sensitivity: 0.35 ± 0.01, AUCROC: 0.64 ± 0.08), the MRI-only deep ensemble (sensitivity: 0.60 ± 0.09, AUCROC: 0.68 ± 0.06), and the 3-channel ensemble without clinical feature fusion (sensitivity: 0.78 ± 0.08, AUCROC: 0.84 ± 0.03).

CONCLUSIONS

Facilitated by the spherical image projection method, a deep ensemble model incorporating Dmap and clinical variables demonstrated excellent performance in predicting BM post-SRS local failure. Our novel approach could improve other radiation therapy outcome models and warrants further evaluation.

CT Cadaveric dataset for Radiomics features stability assessment in lumbar vertebrae

Radiomics features (RFs) studies have showed limitations in the reproducibility of RFs in different acquisition settings. To date, reproducibility studies using CT images mainly rely on phantoms, due to the harness of patient exposure to X-rays. The provided CadAIver dataset has the aims of evaluating how CT scanner parameters effect radiomics features on cadaveric donor. The dataset comprises 112 unique CT acquisitions of a cadaveric truck acquired on 3 different CT scanners varying KV, mA, field-of-view, and reconstruction kernel settings. Technical validation of the CadAIver dataset comprises a comprehensive univariate and multivariate GLM approach to assess stability of each RFs extracted from lumbar vertebrae. The complete dataset is publicly available to be applied for future research in the RFs field, and could foster the creation of a collaborative open CT image database to increase the sample size, the range of available scanners, and the available body districts.

Investigating the role of imaging factors in the variability of CT-based texture analysis metrics

OBJECTIVE

This study assesses the robustness of first-order radiomic texture features namely interquartile range (IQR), coefficient of variation (CV) and standard deviation (SD) derived from computed tomography (CT) images by varying dose, reconstruction algorithms and slice thickness using scans of a uniform water phantom, a commercial anthropomorphic liver phantom, and a human liver in-vivo.

MATERIALS AND METHODS

Scans were acquired on a 16 cm detector GE Revolution Apex Edition CT scanner with variations across three different nominal slice thicknesses: 0.625, 1.25, and 2.5 mm, three different dose levels: CTDIvol of 13.86 mGy for the standard dose, 40% reduced dose and 60% reduced dose and two different reconstruction algorithms: a deep learning image reconstruction (DLIR-high) algorithm and a hybrid iterative reconstruction (IR) algorithm ASiR-V50% (AV50) were explored, varying one at a time. To assess the effect of non-linear modifications of images by AV50 and DLIR-high, images of the water phantom were also reconstructed using filtered back projection (FBP). Quantitative measures of IQR, CV and SD were extracted from twelve pre-selected, circular (1 cm diameter) regions of interest (ROIs) capturing different texture patterns across all scans.

RESULTS

Across all scans, imaging, and reconstruction settings, CV, IQR and SD were observed to increase with reduction in dose and slice thickness. An exception to this observation was found when using FBP reconstruction. Lower values of CV, IQR and SD were observed in DLIR-high reconstructions compared to AV50 and FBP. The Poisson statistics were more stringently noted in FBP than DLIR-high and AV50, due to the non-linear nature of the latter two algorithms.

CONCLUSION

Variation in image noise due to dose reduction algorithms, tube current, and slice thickness show a consistent trend across phantom and patient scans. Prospective evaluation across multiple centers, scanners and imaging protocols is needed for establishing quality assurance standards of radiomics.

Technical and clinical considerations of a physical liver phantom for CT radiomics analysis

OBJECTIVE

This study identifies key characteristics to help build a physical liver computed tomography (CT) phantom for radiomics harmonization; particularly, the higher-order texture metrics.

MATERIALS AND METHODS

CT scans of a radiomics phantom comprising of 18 novel 3D printed inserts with varying size, shape, and material combinations were acquired on a 64-slice CT scanner (Brilliance 64, Philips Healthcare). The images were acquired at 120 kV, 250 mAs, CTDIvol of 16.36 mGy, 2 mm slice thickness, and iterative noise-reduction reconstruction (iDose, Philips Healthcare, Andover, MA). Radiomics analysis was performed using the Cancer Imaging Phenomics Toolkit (CaPTk), following automated segmentation of 3D regions of interest (ROI) of the 18 inserts. The findings were compared to three additional ROI obtained of an anthropomorphic liver phantom, a patient liver CT scan, and a water phantom, at comparable imaging settings. Percentage difference in radiomic metrics values between phantom and tissue was used to assess the biological equivalency and <10% was used to claim equivalent.

RESULTS

The HU for all 18 ROI from the phantom ranged from -30 to 120 which is within clinically observed HU range of the liver, showing that our phantom material (T3-6B) is representative of biological CT tissue densities (liver) with >50% radiomic features having <10% difference from liver tissue. Based on the assessment of the Neighborhood Gray Tone Difference Matrix (NGTDM) metrics it is evident that the water phantom ROI show extreme values compared to the ROIs from the phantom. This result may further reinforce the difference between a structureless quantity such as water HU values and tissue HU values found in liver.

CONCLUSION

The 3-D printed patterns of the constructed radiomics phantom cover a wide span of liver tissue textures seen in CT images. Using our results, texture metrics can be selectively harmonized to establish clinically relevant and reliable radiomics panels.

Protocol selection formalism for minimizing detectable differences in morphological radiomics features of lung lesions in repeated CT acquisitions

Background

The accuracy of morphological radiomic features (MRFs) can be affected by various acquisition settings and imaging conditions. To ensure that clinically irrelevant changes do not reduce sensitivity to capture the radiomics changes between successive acquisitions, it is essential to determine the optimal imaging systems and protocols to use.

Purpose

The main goal of our study was to optimize CT protocols and minimize the minimum detectable difference (MDD) in successive acquisitions of MRFs.

Method

MDDs were derived based on the previous research involving 15 realizations of nodule models at two different sizes. Our study involved simulations of two consecutive acquisitions using 297 different imaging conditions, representing variations in scanners' reconstruction kernels, dose levels, and slice thicknesses. Parametric polynomial models were developed to establish correlations between imaging system characteristics, lesion size, and MDDs. Additionally, polynomial models were used to model the correlation of the imaging system parameters. Optimization problems were formulated for each MRF to minimize the approximated function. Feature importance was determined for each MRF through permutation feature analysis. The proposed method was compared to the recommended guidelines by the quantitative imaging biomarkers alliance (QIBA).

Results

The feature importance analysis showed that lesion size is the most influential parameter to estimate the MDDs in most of the MRFs. Our study revealed that thinner slices and higher doses had a measurable impact on reducing the MDDs. Higher spatial resolution and lower noise magnitude were identified as the most suitable or noninferior acquisition settings. Compared to QIBA, the proposed protocol selection guideline demonstrated a reduced coefficient of variation, with values decreasing from 1.49 to 1.11 for large lesions and from 1.68 to 1.12 for small lesions.

Conclusion

The protocol optimization framework provides means to assess and optimize protocols to minimize the MDD to increase the sensitivity of the measurements in lung cancer screening.

Assessing the robustness of a machine-learning model for early detection of pancreatic adenocarcinoma (PDA): evaluating resilience to variations in image acquisition and radiomics workflow using image perturbation methods

PURPOSE

To evaluate robustness of a radiomics-based support vector machine (SVM) model for detection of visually occult PDA on pre-diagnostic CTs by simulating common variations in image acquisition and radiomics workflow using image perturbation methods.

METHODS

Eighteen algorithmically generated-perturbations, which simulated variations in image noise levels (σ, 2σ, 3σ, 5σ), image rotation [both CT image and the corresponding pancreas segmentation mask by 45° and 90° in axial plane], voxel resampling (isotropic and anisotropic), gray-level discretization [bin width (BW) 32 and 64)], and pancreas segmentation (sequential erosions by 3, 4, 6, and 8 pixels and dilations by 3, 4, and 6 pixels from the boundary), were introduced to the original (unperturbed) test subset (n = 128; 45 pre-diagnostic CTs, 83 control CTs with normal pancreas). Radiomic features were extracted from pancreas masks of these additional test subsets, and the model's performance was compared vis-a-vis the unperturbed test subset.

RESULTS

The model correctly classified 43 out of 45 pre-diagnostic CTs and 75 out of 83 control CTs in the unperturbed test subset, achieving 92.2% accuracy and 0.98 AUC. Model's performance was unaffected by a three-fold increase in noise level except for sensitivity declining to 80% at 3σ (p = 0.02). Performance remained comparable vis-a-vis the unperturbed test subset despite variations in image rotation (p = 0.99), voxel resampling (p = 0.25-0.31), change in gray-level BW to 32 (p = 0.31-0.99), and erosions/dilations up to 4 pixels from the pancreas boundary (p = 0.12-0.34).

CONCLUSION

The model's high performance for detection of visually occult PDA was robust within a broad range of clinically relevant variations in image acquisition and radiomics workflow.

Comparative analysis of radiomics and deep-learning algorithms for survival prediction in hepatocellular carcinoma

To examine the comparative robustness of computed tomography (CT)-based conventional radiomics and deep-learning convolutional neural networks (CNN) to predict overall survival (OS) in HCC patients. Retrospectively, 114 HCC patients with pretherapeutic CT of the liver were randomized into a development (n = 85) and a validation (n = 29) cohort, including patients of all tumor stages and several applied therapies. In addition to clinical parameters, image annotations of the liver parenchyma and of tumor findings on CT were available. Cox-regression based on radiomics features and CNN models were established and combined with clinical parameters to predict OS. Model performance was assessed using the concordance index (C-index). Log-rank tests were used to test model-based patient stratification into high/low-risk groups. The clinical Cox-regression model achieved the best validation performance for OS (C-index [95% confidence interval (CI)] 0.74 [0.57-0.86]) with a significant difference between the risk groups (p = 0.03). In image analysis, the CNN models (lowest C-index [CI] 0.63 [0.39-0.83]; highest C-index [CI] 0.71 [0.49-0.88]) were superior to the corresponding radiomics models (lowest C-index [CI] 0.51 [0.30-0.73]; highest C-index [CI] 0.66 [0.48-0.79]). A significant risk stratification was not possible (p > 0.05). Under clinical conditions, CNN-algorithms demonstrate superior prognostic potential to predict OS in HCC patients compared to conventional radiomics approaches and could therefore provide important information in the clinical setting, especially when clinical data is limited.

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.

Radiomics in the characterization of lipid-poor adrenal adenomas at unenhanced CT: time to look beyond usual density metrics

OBJECTIVES

In this study, we developed a radiomic signature for the classification of benign lipid-poor adenomas, which may potentially help clinicians limit the number of unnecessary investigations in clinical practice. Indeterminate adrenal lesions of benign and malignant nature may exhibit different values of key radiomics features.

METHODS

Patients who had available histopathology reports and a non-contrast-enhanced CT scan were included in the study. Radiomics feature extraction was done after the adrenal lesions were contoured. The primary feature selection and prediction performance scores were calculated using the least absolute shrinkage and selection operator (LASSO). To eliminate redundancy, the best-performing features were further examined using the Pearson correlation coefficient, and new predictive models were created.

RESULTS

This investigation covered 50 lesions in 48 patients. After LASSO-based radiomics feature selection, the test dataset's 30 iterations of logistic regression models produced an average performance of 0.72. The model with the best performance, made up of 13 radiomics features, had an AUC of 0.99 in the training phase and 1.00 in the test phase. The number of features was lowered to 5 after performing Pearson's correlation to prevent overfitting. The final radiomic signature trained a number of machine learning classifiers, with an average AUC of 0.93.

CONCLUSIONS

Including more radiomics features in the identification of adenomas may improve the accuracy of NECT and reduce the need for additional imaging procedures and clinical workup, according to this and other recent radiomics studies that have clear points of contact with current clinical practice.

CLINICAL RELEVANCE STATEMENT

The study developed a radiomic signature using unenhanced CT scans for classifying lipid-poor adenomas, potentially reducing unnecessary investigations that scored a final accuracy of 93%.

KEY POINTS

• Radiomics has potential for differentiating lipid-poor adenomas and avoiding unnecessary further investigations. • Quadratic mean, strength, maximum 3D diameter, volume density, and area density are promising predictors for adenomas. • Radiomics models reach high performance with average AUC of 0.95 in the training phase and 0.72 in the test phase.

Impact of deep learning on radiologists and radiology residents in detecting breast cancer on CT: a cross-vendor test study

AIM

To investigate the effect of deep learning on the diagnostic performance of radiologists and radiology residents in detecting breast cancers on computed tomography (CT).

MATERIALS AND METHODS

In this retrospective study, patients undergoing contrast-enhanced chest CT between January 2010 and December 2020 using equipment from two vendors were included. Patients with confirmed breast cancer were categorised as the training (n=201) and validation (n=26) group and the testing group (n=30) using processed CT images from either vendor. The trained deep-learning model was applied to test group patients with (30 females; mean age = 59.2 ± 15.8 years) and without (19 males, 21 females; mean age = 64 ± 15.9 years) breast cancer. Image-based diagnostic performance of the deep-learning model was evaluated with the area under the receiver operating characteristic curve (AUC). Two radiologists and three radiology residents were asked to detect malignant lesions by recording a four-point diagnostic confidence score before and after referring to the result from the deep-learning model, and their diagnostic performance was evaluated using jackknife alternative free-response receiver operating characteristic analysis by calculating the figure of merit (FOM).

RESULTS

The AUCs of the trained deep-learning model on the validation and test data were 0.976 and 0.967, respectively. After referencing with the result of the deep learning model, the FOMs of readers significantly improved (reader 1/2/3/4/5: from 0.933/0.962/0.883/0.944/0.867 to 0.958/0.968/0.917/0.947/0.900; p=0.038).

CONCLUSION

Deep learning can help radiologists and radiology residents detect breast cancer on CT.

Epicardial adipose tissue, metabolic disorders, and cardiovascular diseases: recent advances classified by research methodologies

Epicardial adipose tissue (EAT) is located between the myocardium and visceral pericardium. The unique anatomy and physiology of the EAT determines its great potential in locally influencing adjacent tissues such as the myocardium and coronary arteries. Classified by research methodologies, this study reviews the latest research progress on the role of EAT in cardiovascular diseases (CVDs), particularly in patients with metabolic disorders. Studies based on imaging techniques demonstrated that increased EAT amount in patients with metabolic disorders is associated with higher risk of CVDs and increased mortality. Then, in-depth profiling studies indicate that remodeled EAT may serve as a local mediator of the deleterious effects of cardiometabolic conditions and plays a crucial role in CVDs. Further, in vitro coculture studies provided preliminary evidence that the paracrine effect of remodeled EAT on adjacent cardiomyocytes can promote the occurrence and progression of CVDs. Considering the important role of EAT in CVDs, targeting EAT might be a potential strategy to reduce cardiovascular risks. Several interventions have been proved effective in reducing EAT amount. Our review provides valuable insights of the relationship between EAT, metabolic disorders, and CVDs, as well as an overview of the methodological constructs of EAT-related studies.

CT-Based Radiomics Predicts the Malignancy of Pulmonary Nodules: A Systematic Review and Meta-Analysis

RATIONALE AND OBJECTIVES

More pulmonary nodules (PNs) have been detected with the wide application of computed tomography (CT) in lung cancer screening. Radiomics is a noninvasive approach to predict the malignancy of PNs. We aimed to systematically evaluate the methodological quality of the eligible studies regarding CT-based radiomics models in predicting the malignancy of PNs and evaluate the model performance of the available studies.

MATERIALS AND METHODS

PubMed, Embase, and Web of Science were searched to retrieve relevant studies. The methodological quality of the included studies was assessed using the Radiomics Quality Score (RQS) and Prediction model Risk of Bias Assessment Tool. A meta-analysis was conducted to evaluate the performance of CT-based radiomics model. Meta-regression and subgroup analyses were employed to investigate the source of heterogeneity.

RESULTS

In total, 49 studies were eligible for qualitative analysis and 27 studies were included in quantitative synthesis. The median RQS of 49 studies was 13 (range -2 to 20). The overall risk of bias was found to be high, and the overall applicability was of low concern in all included studies. The pooled sensitivity, specificity, and diagnostic odds ratio were 0.86 95% confidence interval (CI): 0.79-0.91, 0.84 95% CI: 0.78-0.88, and 31.55 95% CI: 21.31-46.70, respectively. The overall area under the curve was 0.91 95% CI: 0.89-0.94. Meta-regression showed the type of PNs on heterogeneity. CT-based radiomics models performed better in studies including only solid PNs.

CONCLUSION

CT-based radiomics models exhibited excellent diagnostic performance in predicting the malignancy of PNs. Prospective, large sample size, and well-devised studies are desired to verify the prediction capabilities of CT-based radiomics model.

A novel CT-responsive hydrogel for the construction of an organ simulation phantom for the repeatability and stability study of radiomic features

Radiomic features have demonstrated reliable outcomes in tumor grading and detecting precancerous lesions in medical imaging analysis. However, the repeatability and stability of these features have faced criticism. In this study, we aim to enhance the repeatability and stability of radiomic features by introducing a novel CT-responsive hydrogel material. The newly developed CT-responsive hydrogel, mineralized by in situ metal ions, exhibits exceptional repeatability, stability, and uniformity. Moreover, by adjusting the concentration of metal ions, it achieves remarkable CT similarity comparable to that of human organs on CT scans. To create a phantom, the hydrogel was molded into a universal model, displaying controllable CT values ranging from 53 HU to 58 HU, akin to human liver tissue. Subsequently, 1218 radiomic features were extracted from the CT-responsive hydrogel organ simulation phantom. Impressively, 85-97.2% of the extracted features exhibited good repeatability and stability during coefficient of variability analysis. This finding emphasizes the potential of CT-responsive hydrogel in consistently extracting the same features, providing a novel approach to address the issue of repeatability in radiomic features.

Role of Noncontrast Computed Tomography Parameters in Predicting the Outcome of Extracorporeal Shock Wave Lithotripsy for Upper Urinary Stones Cases: A Meta-analysis

RATIONALE

Extracorporeal shock wave lithotripsy (ESWL) is widely considered the primary approach for managing urinary tract stones. This study aimed to assess the predictive factors associated with non-contrast computed tomography (NCCT)-based parameters of upper urinary stones in relation to the outcomes of ESWL.

MATERIALS AND METHODS

A systematic search was conducted in PubMed, ScienceDirect, Web of Science, and Cochrane Library to identify all relevant studies published up to June 3, 2023. Several NCCT-based parameters to predict ESWL outcomes, comprised of mean stone density (MSD), skin-to-stone distance (SSD), and stone size, were extracted and analyzed using Review Manager software.

RESULTS

Out of 979 publications screened, a total of 39 publications, involving 7869 patients, were enrolled in the analysis. The pooled estimate demonstrated significant differences between MSD, and stone size between successful and failure of stone fragmentation groups, in which lower values of these parameters are associated with successful ESWL outcomes.

CONCLUSION

The results from the current study suggested that lower NCCT parameters, notably MSD, SSD, and stone size, are significantly associated with successful ESWL outcome. However, additional large-scale prospective studies are required to utilize these parameters effectively, and the optimal cutoff value should be determined.

Optimal batch determination for improved harmonization and prognostication of multi-center PET/CT radiomics feature in head and neck cancer

Objective. To determine the optimal approach for identifying and mitigating batch effects in PET/CT radiomics features, and further improve the prognosis of patients with head and neck cancer (HNC), this study investigated the performance of three batch harmonization methods.Approach. Unsupervised harmonization identified the batch labels by K-means clustering. Supervised harmonization regarding the image acquisition factors (center, manufacturer, scanner, filter kernel) as known/given batch labels, and Combat harmonization was then implemented separately and sequentially based on the batch labels, i.e. harmonizing features among batches determined by each factor individually or harmonizing features among batches determined by multiple factors successively. Extensive experiments were conducted to predict overall survival (OS) on public PET/CT datasets that contain 800 patients from 9 centers.Main results. In the external validation cohort, results show that compared to original models without harmonization, Combat harmonization would be beneficial in OS prediction with C-index of 0.687-0.740 versus 0.684-0.767. Supervised harmonization slightly outperformed unsupervised harmonization in all models (C-index: 0.692-0.767 versus 0.684-0.750). Separate harmonization outperformed sequential harmonization in CT_m+clinic and CT_cm+clinic models with C-index of 0.752 and 0.722, respectively, while sequential harmonization involved clinical features in PET_rs+clinic model further improving the performance and achieving the highest C-index of 0.767.Significance. Optimal batch determination especially sequential harmonization for Combat holds the potential to improve the prognostic power of radiomics model in multi-center HNC dataset with PET/CT imaging.

A Radiomics Approach to Identify Immunologically Active Tumor in Patients with Head and Neck Squamous Cell Carcinomas

BACKGROUND

We recently developed a gene-expression-based HOT score to identify the hot/cold phenotype of head and neck squamous cell carcinomas (HNSCCs), which is associated with the response to immunotherapy. Our goal was to determine whether radiomic profiling from computed tomography (CT) scans can distinguish hot and cold HNSCC.

METHOD

We included 113 patients from The Cancer Genome Atlas (TCGA) and 20 patients from the Groupe Hospitalier Pitié-Salpêtrière (GHPS) with HNSCC, all with available pre-treatment CT scans. The hot/cold phenotype was computed for all patients using the HOT score. The IBEX software (version 4.11.9, accessed on 30 march 2020) was used to extract radiomic features from the delineated tumor region in both datasets, and the intraclass correlation coefficient (ICC) was computed to select robust features. Machine learning classifier models were trained and tested in the TCGA dataset and validated using the area under the receiver operator characteristic curve (AUC) in the GHPS cohort.

RESULTS

A total of 144 radiomic features with an ICC >0.9 was selected. An XGBoost model including these selected features showed the best performance prediction of the hot/cold phenotype with AUC = 0.86 in the GHPS validation dataset.

CONCLUSIONS AND RELEVANCE

We identified a relevant radiomic model to capture the overall hot/cold phenotype of HNSCC. This non-invasive approach could help with the identification of patients with HNSCC who may benefit from immunotherapy.

Comparing effectiveness of image perturbation and test retest imaging in improving radiomic model reliability

Image perturbation is a promising technique to assess radiomic feature repeatability, but whether it can achieve the same effect as test-retest imaging on model reliability is unknown. This study aimed to compare radiomic model reliability based on repeatable features determined by the two methods using four different classifiers. A 191-patient public breast cancer dataset with 71 test-retest scans was used with pre-determined 117 training and 74 testing samples. We collected apparent diffusion coefficient images and manual tumor segmentations for radiomic feature extraction. Random translations, rotations, and contour randomizations were performed on the training images, and intra-class correlation coefficient (ICC) was used to filter high repeatable features. We evaluated model reliability in both internal generalizability and robustness, which were quantified by training and testing AUC and prediction ICC. Higher testing performance was found at higher feature ICC thresholds, but it dropped significantly at ICC = 0.95 for the test-retest model. Similar optimal reliability can be achieved with testing AUC = 0.7-0.8 and prediction ICC > 0.9 at the ICC threshold of 0.9. It is recommended to include feature repeatability analysis using image perturbation in any radiomic study when test-retest is not feasible, but care should be taken when deciding the optimal feature repeatability criteria.

Photon-counting detector CT-based virtual monoenergetic reconstructions: repeatability and reproducibility of radiomics features of an organic phantom and human myocardium

BACKGROUND

Photon-counting detector computed tomography (PCD-CT) may influence imaging characteristics for various clinical conditions due to higher signal and contrast-to-noise ratio in virtual monoenergetic images (VMI). Radiomics analysis relies on quantification of image characteristics. We evaluated the impact of different VMI reconstructions on radiomic features in in vitro and in vivo PCD-CT datasets.

METHODS

An organic phantom consisting of twelve samples (four oranges, four onions, and four apples) was scanned five times. Twenty-three patients who had undergone coronary computed tomography angiography on a first generation PCD-CT system with the same image acquisitions were analyzed. VMIs were reconstructed at 6 keV levels (40, 55, 70, 90, 120, and 190 keV). The phantoms and the patients' left ventricular myocardium (LVM) were segmented for all reconstructions. Ninety-three original radiomic features were extracted. Repeatability and reproducibility were evaluated through intraclass correlations coefficient (ICC) and post hoc paired samples ANOVA t test.

RESULTS

There was excellent repeatability for radiomic features in phantom scans (all ICC = 1.00). Among all VMIs, 36/93 radiomic features (38.7%) in apples, 28/93 (30.1%) in oranges, and 33/93 (35.5%) in onions were not significantly different. For LVM, the percentage of stable features was high between VMIs ≥ 90 keV (90 versus 120 keV, 77.4%; 90 versus 190 keV, 83.9%; 120 versus 190 keV, 89.3%), while comparison to lower VMI levels led to fewer reproducible features (40 versus 55 keV, 8.6%).

CONCLUSIONS

VMI levels influence the stability of radiomic features in an organic phantom and patients' LVM; stability decreases considerably below 90 keV.

RELEVANCE STATEMENT

Spectral reconstructions significantly influence radiomic features in vitro and in vivo, necessitating standardization and careful attention to these reconstruction parameters before clinical implementation.

KEY POINTS

• Radiomic features have an excellent repeatability within the same PCD-CT acquisition and reconstruction. • Differences in VMI lead to decreased reproducibility for radiomic features. • VMI ≥ 90 keV increased the reproducibility of the radiomic features.

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.

Can Whole-Body Baseline CT Radiomics Add Information to the Prediction of Best Response, Progression-Free Survival, and Overall Survival of Stage IV Melanoma Patients Receiving First-Line Targeted Therapy: A Retrospective Register Study

BACKGROUND

The aim of this study was to investigate whether the combination of radiomics and clinical parameters in a machine-learning model offers additive information compared with the use of only clinical parameters in predicting the best response, progression-free survival after six months, as well as overall survival after six and twelve months in patients with stage IV malignant melanoma undergoing first-line targeted therapy.

METHODS

A baseline machine-learning model using clinical variables (demographic parameters and tumor markers) was compared with an extended model using clinical variables and radiomic features of the whole tumor burden, utilizing repeated five-fold cross-validation. Baseline CTs of 91 stage IV malignant melanoma patients, all treated in the same university hospital, were identified in the Central Malignant Melanoma Registry and all metastases were volumetrically segmented (n = 4727).

RESULTS

Compared with the baseline model, the extended radiomics model did not add significantly more information to the best-response prediction (AUC [95% CI] 0.548 (0.188, 0.808) vs. 0.487 (0.139, 0.743)), the prediction of PFS after six months (AUC [95% CI] 0.699 (0.436, 0.958) vs. 0.604 (0.373, 0.867)), or the overall survival prediction after six and twelve months (AUC [95% CI] 0.685 (0.188, 0.967) vs. 0.766 (0.433, 1.000) and AUC [95% CI] 0.554 (0.163, 0.781) vs. 0.616 (0.271, 1.000), respectively).

CONCLUSIONS

The results showed no additional value of baseline whole-body CT radiomics for best-response prediction, progression-free survival prediction for six months, or six-month and twelve-month overall survival prediction for stage IV melanoma patients receiving first-line targeted therapy. These results need to be validated in a larger cohort.

Imaging of human papilloma virus (HPV) related oropharynx tumour: what we know to date

The tumours of head and neck district are around 3% of all malignancies and squamous cell carcinoma is the most frequent histotype, with rapid increase during the last two decades because of the increment of the infection due to human papilloma virus (HPV). Even if the gold standard for the diagnosis is histological examination, including the detection of viral DNA and transcription products, imaging plays a fundamental role in the detection and staging of HPV + tumours, in order to assess the primary tumour, to establish the extent of disease and for follow-up. The main diagnostic tools are Computed Tomography (CT), Positron Emission Tomography-Computed Tomography (PET-CT) and Magnetic Resonance Imaging (MRI), but also Ultrasound (US) and the use of innovative techniques such as Radiomics have an important role. Aim of our review is to illustrate the main imaging features of HPV + tumours of the oropharynx, in US, CT and MRI imaging. In particular, we will outline the main limitations and strengths of the various imaging techniques, the main uses in the diagnosis, staging and follow-up of disease and the fundamental differential diagnoses of this type of tumour. Finally, we will focus on the innovative technique of texture analysis, which is increasingly gaining importance as a diagnostic tool in aid of the radiologist.

Development and validation of a deep learning radiomics nomogram for preoperatively differentiating thymic epithelial tumor histologic subtypes

OBJECTIVES

Using contrast-enhanced computed tomography (CECT) and deep learning technology to develop a deep learning radiomics nomogram (DLRN) to preoperative predict risk status of patients with thymic epithelial tumors (TETs).

METHODS

Between October 2008 and May 2020, 257 consecutive patients with surgically and pathologically confirmed TETs were enrolled from three medical centers. We extracted deep learning features from all lesions using a transformer-based convolutional neural network and created a deep learning signature (DLS) using selector operator regression and least absolute shrinkage. The predictive capability of a DLRN incorporating clinical characteristics, subjective CT findings and DLS was evaluated by the area under the curve (AUC) of a receiver operating characteristic curve.

RESULTS

To construct a DLS, 25 deep learning features with non-zero coefficients were selected from 116 low-risk TETs (subtypes A, AB, and B1) and 141 high-risk TETs (subtypes B2, B3, and C). The combination of subjective CT features such as infiltration and DLS demonstrated the best performance in differentiating TETs risk status. The AUCs in the training, internal validation, external validation 1 and 2 cohorts were 0.959 (95% confidence interval [CI]: 0.924-0.993), 0.868 (95% CI: 0.765-0.970), 0.846 (95% CI: 0.750-0.942), and 0.846 (95% CI: 0.735-0.957), respectively. The DeLong test and decision in curve analysis revealed that the DLRN was the most predictive and clinically useful model.

CONCLUSIONS

The DLRN comprised of CECT-derived DLS and subjective CT findings showed a high performance in predicting risk status of patients with TETs.

CLINICAL RELEVANCE STATEMENT

Accurate risk status assessment of thymic epithelial tumors (TETs) may aid in determining whether preoperative neoadjuvant treatment is necessary. A deep learning radiomics nomogram incorporating enhancement CT-based deep learning features, clinical characteristics, and subjective CT findings has the potential to predict the histologic subtypes of TETs, which can facilitate decision-making and personalized therapy in clinical practice.

KEY POINTS

• A non-invasive diagnostic method that can predict the pathological risk status may be useful for pretreatment stratification and prognostic evaluation in TET patients. • DLRN demonstrated superior performance in differentiating the risk status of TETs when compared to the deep learning signature, radiomics signature, or clinical model. • The DeLong test and decision in curve analysis revealed that the DLRN was the most predictive and clinically useful in differentiating the risk status of TETs.

CTFlow: Mitigating Effects of Computed Tomography Acquisition and Reconstruction with Normalizing Flows

Mitigating the effects of image appearance due to variations in computed tomography (CT) acquisition and reconstruction parameters is a challenging inverse problem. We present CTFlow, a normalizing flows-based method for harmonizing CT scans acquired and reconstructed using different doses and kernels to a target scan. Unlike existing state-of-the-art image harmonization approaches that only generate a single output, flow-based methods learn the explicit conditional density and output the entire spectrum of plausible reconstruction, reflecting the underlying uncertainty of the problem. We demonstrate how normalizing flows reduces variability in image quality and the performance of a machine learning algorithm for lung nodule detection. We evaluate the performance of CTFlow by 1) comparing it with other techniques on a denoising task using the AAPM-Mayo Clinical Low-Dose CT Grand Challenge dataset, and 2) demonstrating consistency in nodule detection performance across 186 real-world low-dose CT chest scans acquired at our institution. CTFlow performs better in the denoising task for both peak signal-to-noise ratio and perceptual quality metrics. Moreover, CTFlow produces more consistent predictions across all dose and kernel conditions than generative adversarial network (GAN)-based image harmonization on a lung nodule detection task. The code is available at https://github.com/hsu-lab/ctflow.

The Use of Intraoperative CT Hounsfield Unit Values for the Assessment of Bone Quality in Patients Undergoing Lumbar Interbody Fusion

STUDY DESIGN

Retrospective Cohort Study.

OBJECTIVE

To evaluate the accuracy of intraoperatively measured computed tomography (CT) Hounsfield unit (HU) values by comparison with preoperative CT HU values and to compare the radiation exposure between preoperative and intraoperative CT scans.

METHODS

HU values of lumbar vertebrae were measured and compared between preoperative and intraoperative CT scans in patients undergoing lumbar interbody fusion. In patient group one, Canon CT scanners were used preoperatively and the AIRO CT scanner was used intraoperatively. In patient group two, Canon CT scanners were used preoperatively and the O-arm Cone Beam CT (CBCT) scanner was used intraoperatively. In a subgroup analysis of patient group one, radiation by means of CT Dose Index (CTDI) was compared between Canon and AIRO CT scanners.

RESULTS

In the first patient group, a total of 250 vertebrae were analysed in 74 patients showing a strong Pearson correlation of >.94 between pre- and intraoperative HU values. Bland-Altman analysis indicated consistency and equivalence with a bias of 3.9 and 95% limits of agreement from -27.17 to 34.97 when comparing all pre- and intraoperative HU values of L1-5. In the second patient group, a total of 27 vertebrae were analysed in 10 patients showing weak Pearson correlation and Bland-Altman analysis indicated no equivalence. CTDI did not differ between Canon and AIRO CT scanners.

CONCLUSION

Correct and reliable CT HU measurement as mandatory key factor for the intraoperative assessment of bone quality and robotic-assisted surgery is feasible with intraoperative AIRO CT imaging without increase of radiation exposure.

DeepNet model empowered cuckoo search algorithm for the effective identification of lung cancer nodules

Introduction

Globally, lung cancer is a highly harmful type of cancer. An efficient diagnosis system can enable pathologists to recognize the type and nature of lung nodules and the mode of therapy to increase the patient's chance of survival. Hence, implementing an automatic and reliable system to segment lung nodules from a computed tomography (CT) image is useful in the medical industry.

Methods

This study develops a novel fully convolutional deep neural network (hereafter called DeepNet) model for segmenting lung nodules from CT scans. This model includes an encoder/decoder network that achieves pixel-wise image segmentation. The encoder network exploits a Visual Geometry Group (VGG-19) model as a base architecture, while the decoder network exploits 16 upsampling and deconvolution modules. The encoder used in this model has a very flexible structural design that can be modified and trained for any resolution based on the size of input scans. The decoder network upsamples and maps the low-resolution attributes of the encoder. Thus, there is a considerable drop in the number of variables used for the learning process as the network recycles the pooling indices of the encoder for segmentation. The Thresholding method and the cuckoo search algorithm determines the most useful features when categorizing cancer nodules.

Results and discussion

The effectiveness of the intended DeepNet model is cautiously assessed on the real-world database known as The Cancer Imaging Archive (TCIA) dataset and its effectiveness is demonstrated by comparing its representation with some other modern segmentation models in terms of selected performance measures. The empirical analysis reveals that DeepNet significantly outperforms other prevalent segmentation algorithms with 0.962 ± 0.023% of volume error, 0.968 ± 0.011 of dice similarity coefficient, 0.856 ± 0.011 of Jaccard similarity index, and 0.045 ± 0.005s average processing time.

Methodological quality of radiomic-based prognostic studies in gastric cancer: a cross-sectional study

Background

Machine learning radiomics models are increasingly being used to predict gastric cancer prognoses. However, the methodological quality of these models has not been evaluated. Therefore, this study aimed to evaluate the methodological quality of radiomics studies in predicting the prognosis of gastric cancer, summarize their methodological characteristics and performance.

Methods

The PubMed and Embase databases were searched for radiomics studies used to predict the prognosis of gastric cancer published in last 5 years. The characteristics of the studies and the performance of the models were extracted from the eligible full texts. The methodological quality, reporting completeness and risk of bias of the included studies were evaluated using the RQS, TRIPOD and PROBAST. The discrimination ability scores of the models were also compared.

Results

Out of 283 identified records, 22 studies met the inclusion criteria. The study endpoints included survival time, treatment response, and recurrence, with reported discriminations ranging between 0.610 and 0.878 in the validation dataset. The mean overall RQS value was 15.32 ± 3.20 (range: 9 to 21). The mean adhered items of the 35 item of TRIPOD checklist was 20.45 ± 1.83. The PROBAST showed all included studies were at high risk of bias.

Conclusion

The current methodological quality of gastric cancer radiomics studies is insufficient. Large and reasonable sample, prospective, multicenter and rigorously designed studies are required to improve the quality of radiomics models for gastric cancer prediction.

Study registration

This protocol was prospectively registered in the Open Science Framework Registry (https://osf.io/ja52b).

Optimization and validation of voxel size-related radiomics variability by combatting batch effect harmonization in pulmonary nodules: a phantom and clinical study

Background

Broad generalization of radiomics-assisted models may be impeded by concerns about variability. This study aimed to evaluate the merit of combatting batch effect (ComBat) harmonization in reducing the variability of voxel size-related radiomics in both phantom and clinical study in comparison with image resampling correction method.

Methods

A pulmonary phantom with 22 different types of nodules was scanned by computed tomography (CT) with different voxel sizes. The variability of voxel size-related radiomics features was evaluated using concordance correlation coefficient (CCC), dynamic range (DR), and intraclass correlation coefficient (ICC). ComBat and image resampling compensation methods were used to reduce variability of voxel size-related radiomics. The percentage of robust radiomics features was compared before and after optimization. Pathologically differential diagnosis of invasive adenocarcinoma (IAC) from adenocarcinoma in situ (AIS) and minimally invasive adenocarcinoma (MIA) (AIS-MIA group) was used for clinical validation in 134 patients.

Results

Before optimization, the number of excellent features in the phantom and clinical data was 26.12% and 32.31%, respectively. The excellent features were increased after image resampling and ComBat correction. For clinical optimization, the effect of the ComBat compensation method was significantly better than that of image resampling, with excellent features reaching 90.96% and poor features only amounting to 4.96%. In addition, the hierarchical clustering analysis showed that the first-order and shape features had better robustness than did texture features. In clinical validation, the area under the curve (AUC) of the testing set was 0.865 after ComBat correction.

Conclusions

The ComBat harmonization can optimize voxel size-related CT radiomics variability in pulmonary nodules more efficiently than image resampling harmonization.

Spatially-aware clustering improves AJCC-8 risk stratification performance in oropharyngeal carcinomas

OBJECTIVE

Evaluate the effectiveness of machine learning tools that incorporate spatial information such as disease location and lymph node metastatic patterns-of-spread, for prediction of survival and toxicity in HPV+ oropharyngeal cancer (OPC).

MATERIALS & METHODS

675 HPV+ OPC patients that were treated at MD Anderson Cancer Center between 2005 and 2013 with curative intent IMRT were retrospectively collected under IRB approval. Risk stratifications incorporating patient radiometric data and lymph node metastasis patterns via an anatomically-adjacent representation with hierarchical clustering were identified. These clusterings were combined into a 3-level patient stratification and included along with other known clinical features in a Cox model for predicting survival outcomes, and logistic regression for toxicity, using independent subsets for training and validation.

RESULTS

Four groups were identified and combined into a 3-level stratification. The inclusion of patient stratifications in predictive models for 5-yr Overall survival (OS), 5-year recurrence free survival, (RFS) and Radiation-associated dysphagia (RAD) consistently improved model performance measured using the area under the curve (AUC). Test set AUC improvements over models with clinical covariates, was 9 % for predicting OS, and 18 % for predicting RFS, and 7 % for predicting RAD. For models with both clinical and AJCC covariates, AUC improvement was 7 %, 9 %, and 2 % for OS, RFS, and RAD, respectively.

CONCLUSION

Including data-driven patient stratifications considerably improve prognosis for survival and toxicity outcomes over the performance achieved by clinical staging and clinical covariates alone. These stratifications generalize well to across cohorts, and sufficient information for reproducing these clusters is included.

The Promise and Future of Radiomics for Personalized Radiotherapy Dosing and Adaptation

Quantitative image analysis, also known as radiomics, aims to analyze large-scale quantitative features extracted from acquired medical images using hand-crafted or machine-engineered feature extraction approaches. Radiomics has great potential for a variety of clinical applications in radiation oncology, an image-rich treatment modality that utilizes computed tomography (CT), magnetic resonance imaging (MRI), and positron emission tomography (PET) for treatment planning, dose calculation, and image guidance. A promising application of radiomics is in predicting treatment outcomes after radiotherapy such as local control and treatment-related toxicity using features extracted from pretreatment and on-treatment images. Based on these individualized predictions of treatment outcomes, radiotherapy dose can be sculpted to meet the specific needs and preferences of each patient. Radiomics can aid in tumor characterization for personalized targeting, especially for identifying high-risk regions within a tumor that cannot be easily discerned based on size or intensity alone. Radiomics-based treatment response prediction can aid in developing personalized fractionation and dose adjustments. In order to make radiomics models more applicable across different institutions with varying scanners and patient populations, further efforts are needed to harmonize and standardize the acquisition protocols by minimizing uncertainties within the imaging data.

Phantom-based radiomics feature test-retest stability analysis on photon-counting detector CT

OBJECTIVES

Radiomics image data analysis offers promising approaches in research but has not been implemented in clinical practice yet, partly due to the instability of many parameters. The aim of this study is to evaluate the stability of radiomics analysis on phantom scans with photon-counting detector CT (PCCT).

METHODS

Photon-counting CT scans of organic phantoms consisting of 4 apples, kiwis, limes, and onions each were performed at 10 mAs, 50 mAs, and 100 mAs with 120-kV tube current. The phantoms were segmented semi-automatically and original radiomics parameters were extracted. This was followed by statistical analysis including concordance correlation coefficients (CCC), intraclass correlation coefficients (ICC), as well as random forest (RF) analysis, and cluster analysis to determine the stable and important parameters.

RESULTS

Seventy-three of the 104 (70%) extracted features showed excellent stability with a CCC value > 0.9 when compared in a test and retest analysis, and 68 features (65.4%) were stable compared to the original in a rescan after repositioning. Between the test scans with different mAs values, 78 (75%) features were rated with excellent stability. Eight radiomics features were identified that had an ICC value greater than 0.75 in at least 3 of 4 groups when comparing the different phantoms in a phantom group. In addition, the RF analysis identified many features that are important for distinguishing the phantom groups.

CONCLUSION

Radiomics analysis using PCCT data provides high feature stability on organic phantoms, which may facilitate the implementation of radiomics analysis likewise in clinical routine.

KEY POINTS

• Radiomics analysis using photon-counting computed tomography provides high feature stability. • Photon-counting computed tomography may pave the way for implementation of radiomics analysis in clinical routine.

Magnetic resonance imaging in naso-oropharyngeal carcinoma: role of texture analysis in the assessment of response to radiochemotherapy, a preliminary study

OBJECTIVE

Identifying MRI texture parameters able to distinguish inflammation, fibrosis, and residual cancer in patients with naso-oropharynx carcinoma after radiochemotherapy (RT-CHT).

MATERIAL AND METHODS

In this single-centre, observational, retrospective study, texture analysis was performed on ADC maps and post-gadolinium T1 images of patients with histological diagnosis of naso-oropharyngeal carcinoma treated with RT-CHT. An initial cohort of 99 patients was selected; 57 of them were later excluded. The final cohort of 42 patients was divided into 3 groups (inflammation, fibrosis, and residual cancer) according to MRI, 18F-FDG-PET/CT performed 3-4 months after RT-CHT, and biopsy. Pre-RT-CHT lesions and the corresponding anatomic area post-RT-CHT were segmented with 3D slicer software from which 107 textural features were derived. T-Student and Wilcoxon signed-rank tests were performed, and features with p-value < 0.01 were considered statistically significant. Cut-off values-obtained by ROC curves-to discriminate post-RT-CHT non-tumoural changes from residual cancer were calculated for the parameters statistically associated to the diseased status at follow-up.

RESULTS

Two features-Energy and Grey Level Non-Uniformity-were statistically significant on T1 images in the comparison between 'positive' (residual cancer) and 'negative' patients (inflammation and fibrosis). Energy was also found to be statistically significant in both patients with fibrosis and residual cancer. Grey Level Non-Uniformity was significant in the differentiation between residual cancer and inflammation. Five features were statistically significant on ADC maps in the differentiation between 'positive' and 'negative' patients. The reduction in values of such features between pre- and post-RT-CHT was correlated with a good response to therapy.

CONCLUSIONS

Texture analysis on post-gadolinium T1 images and ADC maps can differentiate residual cancer from fibrosis and inflammation in early follow-up of naso-oropharyngeal carcinoma treated with RT-CHT.

A Novel Radiogenomics Biomarker for Predicting Treatment Response and Pneumotoxicity From Programmed Cell Death Protein or Ligand-1 Inhibition Immunotherapy in NSCLC

INTRODUCTION

Patient selection for checkpoint inhibitor immunotherapy is currently guided by programmed death-ligand 1 (PD-L1) expression obtained from immunohistochemical staining of tumor tissue samples. This approach is susceptible to limitations resulting from the dynamic and heterogeneous nature of cancer cells and the invasiveness of the tissue sampling procedure. To address these challenges, we developed a novel computed tomography (CT) radiomic-based signature for predicting disease response in patients with NSCLC undergoing programmed cell death protein 1 (PD-1) or PD-L1 checkpoint inhibitor immunotherapy.

METHODS

This retrospective study comprises a total of 194 patients with suitable CT scans out of 340. Using the radiomic features computed from segmented tumors on a discovery set of 85 contrast-enhanced chest CTs of patients diagnosed with having NSCLC and their CD274 count, RNA expression of the protein-encoding gene for PD-L1, as the response vector, we developed a composite radiomic signature, lung cancer immunotherapy-radiomics prediction vector (LCI-RPV). This was validated in two independent testing cohorts of 66 and 43 patients with NSCLC treated with PD-1 or PD-L1 inhibition immunotherapy, respectively.

RESULTS

LCI-RPV predicted PD-L1 positivity in both NSCLC testing cohorts (area under the curve [AUC] = 0.70, 95% confidence interval [CI]: 0.57-0.84 and AUC = 0.70, 95% CI: 0.46-0.94). In one cohort, it also demonstrated good prediction of cases with high PD-L1 expression exceeding key treatment thresholds (>50%: AUC = 0.72, 95% CI: 0.59-0.85 and >90%: AUC = 0.66, 95% CI: 0.45-0.88), the tumor's objective response to treatment at 3 months (AUC = 0.68, 95% CI: 0.52-0.85), and pneumonitis occurrence (AUC = 0.64, 95% CI: 0.48-0.80). LCI-RPV achieved statistically significant stratification of the patients into a high- and low-risk survival group (hazard ratio = 2.26, 95% CI: 1.21-4.24, p = 0.011 and hazard ratio = 2.45, 95% CI: 1.07-5.65, p = 0.035).

CONCLUSIONS

A CT radiomics-based signature developed from response vector CD274 can aid in evaluating patients' suitability for PD-1 or PD-L1 checkpoint inhibitor immunotherapy in NSCLC.

Artificial intelligence-driven radiomics study in cancer: the role of feature engineering and modeling

Modern medicine is reliant on various medical imaging technologies for non-invasively observing patients' anatomy. However, the interpretation of medical images can be highly subjective and dependent on the expertise of clinicians. Moreover, some potentially useful quantitative information in medical images, especially that which is not visible to the naked eye, is often ignored during clinical practice. In contrast, radiomics performs high-throughput feature extraction from medical images, which enables quantitative analysis of medical images and prediction of various clinical endpoints. Studies have reported that radiomics exhibits promising performance in diagnosis and predicting treatment responses and prognosis, demonstrating its potential to be a non-invasive auxiliary tool for personalized medicine. However, radiomics remains in a developmental phase as numerous technical challenges have yet to be solved, especially in feature engineering and statistical modeling. In this review, we introduce the current utility of radiomics by summarizing research on its application in the diagnosis, prognosis, and prediction of treatment responses in patients with cancer. We focus on machine learning approaches, for feature extraction and selection during feature engineering and for imbalanced datasets and multi-modality fusion during statistical modeling. Furthermore, we introduce the stability, reproducibility, and interpretability of features, and the generalizability and interpretability of models. Finally, we offer possible solutions to current challenges in radiomics research.

Robustness of radiomics features of virtual unenhanced and virtual monoenergetic images in dual-energy CT among different imaging platforms and potential role of CT number variability

OBJECTIVES

To evaluate robustness of dual-energy CT (DECT) radiomics features of virtual unenhanced (VUE) image and virtual monoenergetic image (VMI) among different imaging platforms.

METHODS

A phantom with sixteen clinical-relevant densities was scanned on ten DECT platforms with comparable scan parameters. Ninety-four radiomic features were extracted via Pyradiomics from VUE images and VMIs at energy level of 70 keV (VMI_70keV). Test-retest repeatability was assessed by Bland-Altman analysis. Inter-platform reproducibility of VUE images and VMI_70keV was evaluated by coefficient of variation (CV) and quartile coefficient of dispersion (QCD) among platforms, and by intraclass correlation coefficient (ICC) and concordance correlation coefficient (CCC) between platform pairs. The correlation between variability of CT number radiomics reproducibility was estimated.

RESULTS

92.02% and 92.87% of features were repeatable between scan-rescans for VUE images and VMI_70keV, respectively. Among platforms, 11.30% and 28.39% features of VUE images, and 15.16% and 28.99% features of VMI_70keV were with CV < 10% and QCD < 10%. The average percentages of radiomics features with ICC > 0.90 and CCC > 0.90 between platform pairs were 10.00% and 9.86% in VUE images and 11.23% and 11.23% in VMI_70keV. The CT number inter-platform reproducibility using CV and QCD showed negative correlations with percentage of the first-order radiomics features with CV < 10% and QCD < 10%, in both VUE images and VMI_70keV (r² 0.3870-0.6178, all p < 0.001).

CONCLUSIONS

The majority of DECT radiomics features were non-reproducible. The differences in CT number were considered as an indicator of inter-platform DECT radiomics variation. Critical relevance statement: The majority of radiomics features extracted from the VUE images and the VMI70keV were non-reproducible among platforms, while synchronizing energy levels of VMI to reduce the CT number value variability may be a potential way to mitigate radiomics instability.

A multi-kernel and multi-scale learning based deep ensemble model for predicting recurrence of non-small cell lung cancer

Predicting recurrence in patients with non-small cell lung cancer (NSCLC) before treatment is vital for guiding personalized medicine. Deep learning techniques have revolutionized the application of cancer informatics, including lung cancer time-to-event prediction. Most existing convolutional neural network (CNN) models are based on a single two-dimensional (2D) computational tomography (CT) image or three-dimensional (3D) CT volume. However, studies have shown that using multi-scale input and fusing multiple networks provide promising performance. This study proposes a deep learning-based ensemble network for recurrence prediction using a dataset of 530 patients with NSCLC. This network assembles 2D CNN models of various input slices, scales, and convolutional kernels, using a deep learning-based feature fusion model as an ensemble strategy. The proposed framework is uniquely designed to benefit from (i) multiple 2D in-plane slices to provide more information than a single central slice, (ii) multi-scale networks and multi-kernel networks to capture the local and peritumoral features, (iii) ensemble design to integrate features from various inputs and model architectures for final prediction. The ensemble of five 2D-CNN models, three slices, and two multi-kernel networks, using 5 × 5 and 6 × 6 convolutional kernels, achieved the best performance with an accuracy of 69.62%, area under the curve (AUC) of 72.5%, F1 score of 70.12%, and recall of 70.81%. Furthermore, the proposed method achieved competitive results compared with the 2D and 3D-CNN models for cancer outcome prediction in the benchmark studies. Our model is also a potential adjuvant treatment tool for identifying NSCLC patients with a high risk of recurrence.

Repeatability of radiomics studies in colorectal cancer: a systematic review

BACKGROUND

Recently, radiomics has been widely used in colorectal cancer, but many variable factors affect the repeatability of radiomics research. This review aims to analyze the repeatability of radiomics studies in colorectal cancer and to evaluate the current status of radiomics in the field of colorectal cancer.

METHODS

The included studies in this review by searching from the PubMed and Embase databases. Then each study in our review was evaluated using the Radiomics Quality Score (RQS). We analyzed the factors that may affect the repeatability in the radiomics workflow and discussed the repeatability of the included studies.

RESULTS

A total of 188 studies was included in this review, of which only two (2/188, 1.06%) studies controlled the influence of individual factors. In addition, the median score of RQS was 11 (out of 36), range-1 to 27.

CONCLUSIONS

The RQS score was moderately low, and most studies did not consider the repeatability of radiomics features, especially in terms of Intra-individual, scanners, and scanning parameters. To improve the generalization of the radiomics model, it is necessary to further control the variable factors of repeatability.

Pilot Study for the Assessment of the Best Radiomic Features for Bosniak Cyst Classification Using Phantom and Radiologist Inter-Observer Selection

Since the Bosniak cysts classification is highly reader-dependent, automated tools based on radiomics could help in the diagnosis of the lesion. This study is an initial step in the search for radiomic features that may be good classifiers of benign-malignant Bosniak cysts in machine learning models. A CCR phantom was used through five CT scanners. Registration was performed with ARIA software, while Quibim Precision was used for feature extraction. R software was used for the statistical analysis. Robust radiomic features based on repeatability and reproducibility criteria were chosen. Excellent correlation criteria between different radiologists during lesion segmentation were imposed. With the selected features, their classification ability in benignity-malignity terms was assessed. From the phantom study, 25.3% of the features were robust. For the study of inter-observer correlation (ICC) in the segmentation of cystic masses, 82 subjects were prospectively selected, finding 48.4% of the features as excellent regarding concordance. Comparing both datasets, 12 features were established as repeatable, reproducible, and useful for the classification of Bosniak cysts and could serve as initial candidates for the elaboration of a classification model. With those features, the Linear Discriminant Analysis model classified the Bosniak cysts in terms of benignity or malignancy with 88.2% accuracy.

In Vivo Repeatability and Multiscanner Reproducibility of MRI Radiomics Features in Patients With Monoclonal Plasma Cell Disorders: A Prospective Bi-institutional Study

OBJECTIVES

Despite the extensive number of publications in the field of radiomics, radiomics algorithms barely enter large-scale clinical application. Supposedly, the low external generalizability of radiomics models is one of the main reasons, which hinders the translation from research to clinical application. The objectives of this study were to investigate reproducibility of radiomics features (RFs) in vivo under variation of patient positioning, magnetic resonance imaging (MRI) sequence, and MRI scanners, and to identify a subgroup of RFs that shows acceptable reproducibility across all different acquisition scenarios.

MATERIALS AND METHODS

Between November 30, 2020 and February 16, 2021, 55 patients with monoclonal plasma cell disorders were included in this prospective, bi-institutional, single-vendor study. Participants underwent one reference scan at a 1.5 T MRI scanner and several retest scans: once after simple repositioning, once with a second MRI protocol, once at another 1.5 T scanner, and once at a 3 T scanner. Radiomics feature from the bone marrow of the left hip bone were extracted, both from original scans and after different image normalizations. Intraclass correlation coefficient (ICC) was used to assess RF repeatability and reproducibility.

RESULTS

Fifty-five participants (mean age, 59 ± 7 years; 36 men) were enrolled. For T1-weighted images after muscle normalization, in the simple test-retest experiment, 110 (37%) of 295 RFs showed an ICC ≥0.8: 54 (61%) of 89 first-order features (FOFs), 35 (95%) of 37 volume and shape features, and 21 (12%) of 169 texture features (TFs). When the retest was performed with different technical settings, even after muscle normalization, the number of FOF/TF with an ICC ≥0.8 declined to 58/13 for the second protocol, 29/7 for the second 1.5 T scanner, and 49/7 for the 3 T scanner, respectively. Twenty-five (28%) of the 89 FOFs and 6 (4%) of the 169 TFs from muscle-normalized T1-weighted images showed an ICC ≥0.8 throughout all repeatability and reproducibility experiments.

CONCLUSIONS

In vivo, only few RFs are reproducible with different MRI sequences or different MRI scanners, even after application of a simple image normalization. Radiomics features selected by a repeatability experiment only are not necessarily suited to build radiomics models for multicenter clinical application. This study isolated a subset of RFs, which are robust to variations in MRI acquisition observed in scanners from 1 vendor, and therefore are candidates to build reproducible radiomics models for monoclonal plasma cell disorders for multicentric applications, at least when centers are equipped with scanners from this vendor.