Radiomics: a new application from established techniques

Machine Learning-based Differentiation of Benign and Premalignant Colorectal Polyps Detected with CT Colonography in an Asymptomatic Screening Population: A Proof-of-Concept Study

Background CT colonography does not enable definite differentiation between benign and premalignant colorectal polyps. Purpose To perform machine learning-based differentiation of benign and premalignant colorectal polyps detected with CT colonography in an average-risk asymptomatic colorectal cancer screening sample with external validation using radiomics. Materials and Methods In this secondary analysis of a prospective trial, colorectal polyps of all size categories and morphologies were manually segmented on CT colonographic images and were classified as benign (hyperplastic polyp or regular mucosa) or premalignant (adenoma) according to the histopathologic reference standard. Quantitative image features characterizing shape (n = 14), gray level histogram statistics (n = 18), and image texture (n = 68) were extracted from segmentations after applying 22 image filters, resulting in 1906 feature-filter combinations. Based on these features, a random forest classification algorithm was trained to predict the individual polyp character. Diagnostic performance was validated in an external test set. Results The random forest model was fitted using a training set consisting of 107 colorectal polyps in 63 patients (mean age, 63 years ± 8 [standard deviation]; 40 men) comprising 169 segmentations on CT colonographic images. The external test set included 77 polyps in 59 patients comprising 118 segmentations. Random forest analysis yielded an area under the receiver operating characteristic curve of 0.91 (95% CI: 0.85, 0.96), a sensitivity of 82% (65 of 79) (95% CI: 74%, 91%), and a specificity of 85% (33 of 39) (95% CI: 72%, 95%) in the external test set. In two subgroup analyses of the external test set, the area under the receiver operating characteristic curve was 0.87 in the size category of 6-9 mm and 0.90 in the size category of 10 mm or larger. The most important image feature for decision making (relative importance of 3.7%) was quantifying first-order gray level histogram statistics. Conclusion In this proof-of-concept study, machine learning-based image analysis enabled noninvasive differentiation of benign and premalignant colorectal polyps with CT colonography. © RSNA, 2021 Online supplemental material is available for this article.

A Systematic Review of PET Textural Analysis and Radiomics in Cancer

Background: Although many works have supported the utility of PET radiomics, several authors have raised concerns over the robustness and replicability of the results. This study aimed to perform a systematic review on the topic of PET radiomics and the used methodologies. Methods: PubMed was searched up to 15 October 2020. Original research articles based on human data specifying at least one tumor type and PET image were included, excluding those that apply only first-order statistics and those including fewer than 20 patients. Each publication, cancer type, objective and several methodological parameters (number of patients and features, validation approach, among other things) were extracted. Results: A total of 290 studies were included. Lung (28%) and head and neck (24%) were the most studied cancers. The most common objective was prognosis/treatment response (46%), followed by diagnosis/staging (21%), tumor characterization (18%) and technical evaluations (15%). The average number of patients included was 114 (median = 71; range 20–1419), and the average number of high-order features calculated per study was 31 (median = 26, range 1–286). Conclusions: PET radiomics is a promising field, but the number of patients in most publications is insufficient, and very few papers perform in-depth validations. The role of standardization initiatives will be crucial in the upcoming years.

Imaging-Based Staging of Hepatic Fibrosis in Patients with Hepatitis B: A Dynamic Radiomics Model Based on Gd-EOB-DTPA-Enhanced MRI

Accurate grading of liver fibrosis can effectively assess the severity of liver disease and help doctors make an appropriate diagnosis. This study aimed to perform the automatic staging of hepatic fibrosis on patients with hepatitis B, who underwent gadolinium ethoxybenzyl diethylenetriamine pentaacetic acid (Gd-EOB-DTPA)-enhanced magnetic resonance imaging with dynamic radiomics analysis. The proposed dynamic radiomics model combined imaging features from multi-phase dynamic contrast-enhanced (DCE) images and time-domain information. Imaging features were extracted from the deep learning-based segmented liver volume, and time-domain features were further explored to analyze the variation in features during contrast enhancement. Model construction and evaluation were based on a 132-case data set. The proposed model achieved remarkable performance in significant fibrosis (fibrosis stage S1 vs. S2-S4; accuracy (ACC) = 0.875, area under the curve (AUC) = 0.867), advanced fibrosis (S1-S2 vs. S3-S4; ACC = 0.825, AUC = 0.874), and cirrhosis (S1-S3 vs. S4; ACC = 0.850, AUC = 0.900) classifications in the test set. It was more dominant compared with the conventional single-phase or multi-phase DCE-based radiomics models, normalized liver enhancement, and some serological indicators. Time-domain features were found to play an important role in the classification models. The dynamic radiomics model can be applied for highly accurate automatic hepatic fibrosis staging.

Radiomic Machine Learning Classifiers in Spine Bone Tumors: A Multi-Software, Multi-Scanner Study

PURPOSE

Spinal lesion differential diagnosis remains challenging even in MRI. Radiomics and machine learning (ML) have proven useful even in absence of a standardized data mining pipeline. We aimed to assess ML diagnostic performance in spinal lesion differential diagnosis, employing radiomic data extracted by different software.

METHODS

Patients undergoing MRI for a vertebral lesion were retrospectively analyzed (n = 146, 67 males, 79 females; mean age 63 ± 16 years, range 8-89 years) and constituted the train (n = 100) and internal test cohorts (n = 46). Part of the latter had additional prior exams which constituted a multi-scanner, external test cohort (n = 35). Lesions were labeled as benign or malignant (2-label classification), and benign, primary malignant or metastases (3-label classification) for classification analyses. Features extracted via 3D Slicer heterogeneityCAD module (hCAD) and PyRadiomics were independently used to compare different combinations of feature selection methods and ML classifiers (n = 19).

RESULTS

In total, 90 and 1548 features were extracted by hCAD and PyRadiomics, respectively. The best feature selection method-ML algorithm combination was selected by 10 iterations of 10-fold cross-validation in the training data. For the 2-label classification ML obtained 94% accuracy in the internal test cohort, using hCAD data, and 86% in the external one. For the 3-label classification, PyRadiomics data allowed for 80% and 69% accuracy in the internal and external test sets, respectively.

CONCLUSIONS

MRI radiomics combined with ML may be useful in spinal lesion assessment. More robust pre-processing led to better consistency despite scanner and protocol heterogeneity.

Longitudinal acquisition repeatability of MRI radiomics features: An ACR MRI phantom study on two MRI scanners using a 3D T1W TSE sequence

PURPOSE

The purpose of this study was to quantitatively assess the longitudinal acquisition repeatability of MRI radiomics features in a three-dimensional (3D) T1-weighted (T1W) TSE sequence via a well-controlled prospective phantom study.

METHODS

Thirty consecutive daily datasets of an ACR-MRI phantom were acquired on two 1.5T MRI simulators using a 3D T1W TSE sequence. Images were blindly segmented by two observers. Post-acquisition processing was minimized but an intensity discretization (fixed bin size of 25). One hundred and one radiomics features (shape n = 12; first order n = 16; texture n = 73) were extracted. Longitudinal repeatability of each feature was evaluated by Pearson correlation and coefficient of variance (CV_68% ). Interobserver feature value agreement was also quantified using intraclass correlation coefficient (ICC) and Bland-Altman analysis. A most repeatable radiomics feature set on both scanners was determined by feature coefficient of variance (CV_68% <5%), ICC (>0.75), and the ratio of the interobserver difference to the interobserver mean δ<5%.

RESULTS

No trend of radiomics feature value changed with time. Longitudinal feature repeatability CV_68% ranged 0.01-38.60% (mean/median: 12.5%/9.9%), and 0.01-40.47%, (8.49%/7.34%) on the scanners A and B. Shape features exhibited significantly better repeatability than first-order and texture features (all P < 0.01). Significant longitudinal repeatability difference was observed in texture features (P < 0.001) between the two scanners, but not in shape and first-order features (P > 0.30). First-order and texture features had smaller interobserver-dependent variation than acquisition-dependent variation. They also showed good interobserver agreement on both scanners (A:ICC = 0.80 ± 0.23; B:ICC = 0.80 ± 0.22), independent of acquisition repeatability. The repeatable radiomics features in common on both scanners, including 12 shape features, 0 first-order features, and 3 texture features, were determined as the most repeatable MRI radiomics feature set.

CONCLUSIONS

Radiomics features exhibited heterogeneous longitudinal repeatability, while the shape features were the most repeatable, in this phantom study with a 3D T1W TSE acquisition. The most repeatable radiomics feature set derived in this study should be helpful for the selection of reliable radiomics features in the future clinical use.

Lung Nodule Classification Using Biomarkers, Volumetric Radiomics, and 3D CNNs

We present a hybrid algorithm to estimate lung nodule malignancy that combines imaging biomarkers from Radiologist’s annotation with image classification of CT scans. Our algorithm employs a 3D Convolutional Neural Network (CNN) as well as a Random Forest in order to combine CT imagery with biomarker annotation and volumetric radiomic features. We analyze and compare the performance of the algorithm using only imagery, only biomarkers, combined imagery + biomarkers, combined imagery + volumetric radiomic features, and finally the combination of imagery + biomarkers + volumetric features in order to classify the suspicion level of nodule malignancy. The National Cancer Institute (NCI) Lung Image Database Consortium (LIDC) IDRI dataset is used to train and evaluate the classification task. We show that the incorporation of semi-supervised learning by means of K-Nearest-Neighbors (KNN) can increase the available training sample size of the LIDC-IDRI, thereby further improving the accuracy of malignancy estimation of most of the models tested although there is no significant improvement with the use of KNN semi-supervised learning if image classification with CNNs and volumetric features is combined with descriptive biomarkers. Unexpectedly, we also show that a model using image biomarkers alone is more accurate than one that combines biomarkers with volumetric radiomics, 3D CNNs, and semi-supervised learning. We discuss the possibility that this result may be influenced by cognitive bias in LIDC-IDRI because malignancy estimates were recorded by the same radiologist panel as biomarkers, as well as future work to incorporate pathology information over a subset of study participants.

Quantitative Imaging and Radiomics in Multiple Myeloma: A Potential Opportunity?

Multiple Myeloma (MM) is the second most common type of hematological disease and, although it is rare among patients under 40 years of age, its incidence rises in elderly subjects. MM manifestations are usually identified through hyperCalcemia, Renal failure, Anaemia, and lytic Bone lesions (CRAB). In particular, the extent of the bone disease is negatively related to a decreased quality of life in patients and, in general, bone disease in MM increases both morbidity and mortality. The detection of lytic bone lesions on imaging, especially computerized tomography (CT) and Magnetic Resonance Imaging (MRI), is becoming crucial from the clinical viewpoint to separate asymptomatic from symptomatic MM patients and the detection of focal lytic lesions in these imaging data is becoming relevant even when no clinical symptoms are present. Therefore, radiology is pivotal in the staging and accurate management of patients with MM even in early phases of the disease. In this review, we describe the opportunities offered by quantitative imaging and radiomics in multiple myeloma. At the present time there is still high variability in the choice between various imaging methods to study MM patients and high variability in image interpretation with suboptimal agreement among readers even in tertiary centers. Therefore, the potential of medical imaging for patients affected by MM is still to be completely unveiled. In the coming years, new insights to study MM with medical imaging will derive from artificial intelligence (AI) and radiomics usage in different bone lesions and from the wide implementations of quantitative methods to report CT and MRI. Eventually, medical imaging data can be integrated with the patient's outcomes with the purpose of finding radiological biomarkers for predicting the prognostic flow and therapeutic response of the disease.

Feasibility of MRI Radiomics for Predicting KRAS Mutation in Rectal Cancer

The mutation status of KRAS is a significant biomarker in the prognosis of rectal cancer. This study investigated the feasibility of MRI-based radiomics in predicting the mutation status of KRAS with a composite index which could be an important criterion for KRAS mutation in clinical practice. In this retrospective study, a total of 127 patients with rectal cancer were enrolled. The 3D Slicer was used to extract the radiomics features from the MRI images, and sparse support vector machine (SVM) with linear kernel was applied for feature reduction. The radiomics classifier for predicting the KRAS status was then constructed by Linear Discriminant Analysis (LDA) and its performance was evaluated. The composite index was determined with LDA model. Out of 127 rectal cancer subjects, there were 44 KRAS mutation cases and 83 wild cases. A total of 104 radiomics features were extracted, 54 features were filtered by linear SVM with L1-norm regularization and 6 features that had no significant correlations within them were finally selected. The radiomics classifier constructed using the 6 features featured an AUC value of 0.669 (specificity: 0.506; sensitivity: 0.773) with LDA. Furthermore, the composite index (Radscore) had statistically significant difference between the KRAS mutation and wild groups. It is suggested that the MRI-based radiomics has the potential in predicting the KRAS status in patients with rectal cancer, which may enhance the diagnostic value of MRI in rectal cancer.

Screening of COVID-19 based on the extracted radiomics features from chest CT images

BACKGROUND AND OBJECTIVE

Radiomics has been widely used in quantitative analysis of medical images for disease diagnosis and prognosis assessment. The objective of this study is to test a machine-learning (ML) method based on radiomics features extracted from chest CT images for screening COVID-19 cases.

METHODS

The study is carried out on two groups of patients, including 138 patients with confirmed and 140 patients with suspected COVID-19. We focus on distinguishing pneumonia caused by COVID-19 from the suspected cases by segmentation of whole lung volume and extraction of 86 radiomics features. Followed by feature extraction, nine feature-selection procedures are used to identify valuable features. Then, ten ML classifiers are applied to classify and predict COVID-19 cases. Each ML models is trained and tested using a ten-fold cross-validation method. The predictive performance of each ML model is evaluated using the area under the curve (AUC) and accuracy.

RESULTS

The range of accuracy and AUC is from 0.32 (recursive feature elimination [RFE]+Multinomial Naive Bayes [MNB] classifier) to 0.984 (RFE+bagging [BAG], RFE+decision tree [DT] classifiers) and 0.27 (mutual information [MI]+MNB classifier) to 0.997 (RFE+k-nearest neighborhood [KNN] classifier), respectively. There is no direct correlation among the number of the selected features, accuracy, and AUC, however, with changes in the number of the selected features, the accuracy and AUC values will change. Feature selection procedure RFE+BAG classifier and RFE+DT classifier achieve the highest prediction accuracy (accuracy: 0.984), followed by MI+Gaussian Naive Bayes (GNB) and logistic regression (LGR)+DT classifiers (accuracy: 0.976). RFE+KNN classifier as a feature selection procedure achieve the highest AUC (AUC: 0.997), followed by RFE+BAG classifier (AUC: 0.991) and RFE+gradient boosting decision tree (GBDT) classifier (AUC: 0.99).

CONCLUSION

This study demonstrates that the ML model based on RFE+KNN classifier achieves the highest performance to differentiate patients with a confirmed infection caused by COVID-19 from the suspected cases.

Radiomics of hepatocellular carcinoma

The diagnosis of hepatocellular carcinoma relies largely on non-invasive imaging, and is well suited for radiomics analysis. Radiomics is an emerging method for quantification of tumor heterogeneity by mathematically analyzing the spatial distribution and relationships of gray levels in medical images. The published studies on radiomics analysis of HCC provide encouraging data demonstrating potential utility for prediction of tumor biology, molecular profiles, post-therapy response, and outcome. The combination of radiomics data and clinical/laboratory information provides added value in many studies. Radiomics is a multi-step process that requires optimization and standardization, the development of semi-automated or automated segmentation methods, robust data quality control, and refinement of algorithms and modeling approaches for high-throughput data analysis. While radiomics remains largely in the research setting, the strong associations of predictive models and nomograms with certain pathologic, molecular, and immune markers with tumor aggressiveness and patient outcomes, provide great potential for clinical applications to inform optimized treatment strategies and patient prognosis.

Radiomics and Digital Image Texture Analysis in Oncology (Review)

One of the most promising areas of diagnosis and prognosis of diseases is radiomics, a science combining radiology, mathematical modeling, and deep machine learning. The main concept of radiomics is image biomarkers (IBMs), the parameters characterizing various pathological changes and calculated based on the analysis of digital image texture. IBMs are used for quantitative assessment of digital imaging results (CT, MRI, ultrasound, PET). The use of IBMs in the form of "virtual biopsy" is of particular relevance in oncology. The article provides the basic concepts of radiomics identifying the main stages of obtaining IBMs: data collection and preprocessing, tumor segmentation, data detection and extraction, modeling, statistical processing, and data validation. The authors have analyzed the possibilities of using IBMs in oncology, describing the currently known features and advantages of using radiomics and image texture analysis in the diagnosis and prognosis of cancer. The limitations and problems associated with the use of radiomics data are considered. Although the novel effective tool for performing virtual biopsy of human tissue is at the development stage, quite a few projects have already been implemented, and medical software packages for radiomics analysis of digital images have been created.

Stratification of cystic renal masses into benign and potentially malignant: applying machine learning to the bosniak classification

PURPOSE

To create a CT texture-based machine learning algorithm that distinguishes benign from potentially malignant cystic renal masses as defined by the Bosniak Classification version 2019.

METHODS

In this IRB-approved, HIPAA-compliant study, 4,454 adult patients underwent renal mass protocol CT or CT urography from January 2011 to June 2018. Of these, 257 cystic renal masses were included in the final study cohort. Each mass was independently classified using Bosniak version 2019 by three radiologists, resulting in 185 benign (Bosniak I or II) and 72 potentially malignant (Bosniak IIF, III or IV) masses. Six texture features: mean, standard deviation, mean of positive pixels, entropy, skewness, kurtosis were extracted using commercial software TexRAD (Feedback PLC, Cambridge, UK). Random forest (RF), logistic regression (LR), and support vector machine (SVM) machine learning algorithms were implemented to classify cystic renal masses into the two groups and tested with tenfold cross validations.

RESULTS

Higher mean, standard deviation, mean of positive pixels, entropy, skewness were statistically associated with the potentially malignant group (P ≤ 0.0015 each). Sensitivity, specificity, positive predictive value, negative predictive value, and area under curve of RF model was 0.67, 0.91, 0.75, 0.88, 0.88; of LR model was 0.63, 0.93, 0.78, 0.86, 0.90, and of SVM model was 0.56, 0.91, 0.71, 0.84, 0.89, respectively.

CONCLUSION

Three CT texture-based machine learning algorithms demonstrated high discriminatory capability in distinguishing benign from potentially malignant cystic renal masses as defined by the Bosniak Classification version 2019. If validated, CT texture-based machine learning algorithms may help reduce interreader variability when applying the Bosniak classification.

Spatial process decomposition for quantitative imaging biomarkers using multiple images of varying shapes

Quantitative imaging biomarkers (QIB) are extracted from medical images in radiomics for a variety of purposes including noninvasive disease detection, cancer monitoring, and precision medicine. The existing methods for QIB extraction tend to be ad hoc and not reproducible. In this article, a general and flexible statistical approach is proposed for handling up to three-dimensional medical images and reasonably capturing features with respect to specific spatial patterns. In particular, a model-based spatial process decomposition is developed where the random weights are unique to individual patients for component functions common across patients. Model fitting and selection are based on maximum likelihood, while feature extractions are via optimal prediction of the underlying true image. Simulation studies are conducted to investigate the properties of the proposed methodology. For illustration, a cancer image data set is analyzed and QIBs are extracted in association with a clinical endpoint.

Artificial intelligence and hybrid imaging: the best match for personalized medicine in oncology

Artificial intelligence (AI) refers to a field of computer science aimed to perform tasks typically requiring human intelligence. Currently, AI is recognized in the broader technology radar within the five key technologies which emerge for their wide-ranging applications and impact in communities, companies, business, and value chain framework alike. However, AI in medical imaging is at an early phase of development, and there are still hurdles to take related to reliability, user confidence, and adoption. The present narrative review aimed to provide an overview on AI-based approaches (distributed learning, statistical learning, computer-aided diagnosis and detection systems, fully automated image analysis tool, natural language processing) in oncological hybrid medical imaging with respect to clinical tasks (detection, contouring and segmentation, prediction of histology and tumor stage, prediction of mutational status and molecular therapies targets, prediction of treatment response, and outcome). Particularly, AI-based approaches have been briefly described according to their purpose and, finally lung cancer-being one of the most extensively malignancy studied by hybrid medical imaging-has been used as illustrative scenario. Finally, we discussed clinical challenges and open issues including ethics, validation strategies, effective data-sharing methods, regulatory hurdles, educational resources, and strategy to facilitate the interaction among different stakeholders. Some of the major changes in medical imaging will come from the application of AI to workflow and protocols, eventually resulting in improved patient management and quality of life. Overall, several time-consuming tasks could be automatized. Machine learning algorithms and neural networks will permit sophisticated analysis resulting not only in major improvements in disease characterization through imaging, but also in the integration of multiple-omics data (i.e., derived from pathology, genomic, proteomics, and demographics) for multi-dimensional disease featuring. Nevertheless, to accelerate the transition of the theory to practice a sustainable development plan considering the multi-dimensional interactions between professionals, technology, industry, markets, policy, culture, and civil society directed by a mindset which will allow talents to thrive is necessary.

Pancreatic neuroendocrine tumours: spectrum of imaging findings

Pancreatic neuroendocrine tumours (pNETs) are rare and heterogeneous group of neoplasms presenting with a wide variety of symptoms and biological behaviour, from indolent to aggressive ones. pNETs are stratified into functional or non-functional, because of their ability to produce metabolically active hormones. pNETs can be an isolate phenomenon or a part of a hereditary syndrome like von Hippel-Lindau syndrome or neurofibromatosis-1. The incidence has increased in the last years, also because of the improvement of cross-sectional imaging. Computed tomography (CT), magnetic resonance imaging (MRI) and functional imaging are the mainstay imaging modalities used for tumour detection and disease extension assessment, due to easy availability and better contrast/spatial resolution. Radiological imaging plays a fundamental role in detection, characterization and surveillance of pNETs and is involved in almost every stage of patients' management. Moreover, with specific indications and techniques, interventional radiology can also play a role in therapeutic management. Surgery is the treatment of choice, consisting of either partial pancreatectomy or enucleation of the primary tumour. This article reviews the radiologic features of different pNETs as well as imaging mimics, in order to help radiologists to avoid potential pitfalls, to reach the correct diagnosis and to support the multidisciplinary team in establishing the right treatment.

Application of digital mammography-based radiomics in the differentiation of benign and malignant round-like breast tumors and the prediction of molecular subtypes

BACKGROUND

This study aimed to investigate the diagnostic performance of radiomic features based on digital mammography (DM) in the differential diagnosis of benign and malignant round-like (round and oval) solid tumors with circumscribed or obscured margins but without suspicious malignant or benign macrocalcifications and to investigate whether quantitative radiomic features can distinguish triple-negative breast cancer (TNBC) from non-TNBC (NTNBC).

METHODS

This retrospective study included 112 patients with round-like tumors who underwent DM within 20 days preoperatively. Breast masses were segmented manually on the DM images, then radiomic features were extracted. The predictive models were used to distinguish between benign and malignant tumors and to predict TNBC in invasive ductal carcinoma. The receiver operating characteristic curves (ROCs) for these models were obtained for initial DM characteristics, radiomic features to predict malignant tumors and TNBC. The decision curve was obtained to evaluate the clinical usefulness of the model for the prediction of benign or malignant tumors.

RESULTS

The study cohort included 79 patients with pathologically confirmed malignant masses and 33 patients with benign (training cohort: n=79; testing cohort: n=33). A total of 396 features were extracted from the DM images for each patient. The radiomics model for the prediction of malignant tumors achieved an area under the receiver operating characteristic curve (AUC) of 0.88 [95% confidence interval (CI), 0.76-1.00] in the testing cohort; the radiomics model for the prediction of TNBC achieved an AUC of 0.84 (95% CI, 0.73-0.96). In contrast, DM characteristics alone poorly predicted malignant tumors, with the density achieving an AUC 0.69 (95% CI, 0.59-0.79); there was no significant difference in DM characteristics between TNBC and NTNBC (P>0.05, all). The decision curve showed the good clinical usefulness of the model for the prediction of malignant tumors.

CONCLUSIONS

This study showed that DM-based radiomics can accurately discriminate between benign and malignant round-like tumors with circumscribed or obscured margins but without suspicious malignant or benign macrocalcifications. Additionally, it can be used to predict TNBC in invasive ductal carcinoma. DM-based radiomics can aid radiologists in mammogram reading, clinical diagnosis and decision-making.

Correlating Radiomic Features of Heterogeneity on CT with Circulating Tumor DNA in Metastatic Melanoma

Clinical imaging methods, such as computed tomography (CT), are used for routine tumor response monitoring. Imaging can also reveal intratumoral, intermetastatic, and interpatient heterogeneity, which can be quantified using radiomics. Circulating tumor DNA (ctDNA) in the plasma is a sensitive and specific biomarker for response monitoring. Here we evaluated the interrelationship between circulating tumor DNA mutant allele fraction (ctDNAmaf), obtained by targeted amplicon sequencing and shallow whole genome sequencing, and radiomic measurements of CT heterogeneity in patients with stage IV melanoma. ctDNAmaf and radiomic observations were obtained from 15 patients with a total of 70 CT examinations acquired as part of a prospective trial. 26 of 39 radiomic features showed a significant relationship with log(ctDNAmaf). Principal component analysis was used to define a radiomics signature that predicted ctDNAmaf independent of lesion volume. This radiomics signature and serum lactate dehydrogenase were independent predictors of ctDNAmaf. Together, these results suggest that radiomic features and ctDNAmaf may serve as complementary clinical tools for treatment monitoring.

Development and validation of radiomics model built by incorporating machine learning for identifying liver fibrosis and early-stage cirrhosis

BACKGROUND

Liver fibrosis (LF) continues to develop and eventually progresses to cirrhosis. However, LF and early-stage cirrhosis (ESC) can be reversed in some cases, while advanced cirrhosis is almost impossible to cure. Advances in quantitative imaging techniques have made it possible to replace the gold standard biopsy method with non-invasive imaging, such as radiomics. Therefore, the purpose of this study is to develop a radiomics model to identify LF and ESC.

METHODS

Patients with LF (n = 108) and ESC (n = 116) were enrolled in this study. As a control, patients with healthy livers were involved in the study (n = 145). Diffusion-weighted imaging (DWI) data sets with three b-values (0, 400, and 800 s/mm) of enrolled cases were collected in this study. Then, radiomics features were extracted from manually delineated volumes of interest. Two modeling strategies were performed after univariate analysis and feature selection. Finally, an optimal model was determined by the receiver operating characteristic area under the curve (AUC).

RESULTS

The optimal models were built in plan 1. For model 1 in plan 1, the AUCs of the training and validation cohorts were 0.973 (95% confidence interval [CI] 0.946-1.000) and 0.948 (95% CI 0.903-0.993), respectively. For model 2 in plan 1, the AUCs of the training and validation cohorts were 0.944, 95% CI 0.905 to 0.983, and 0.968, 95% CI 0.940 to 0.996, respectively.

CONCLUSIONS

Radiomics analysis of DWI images allows for accurate identification of LF and ESC, and the non-invasive biomarkers extracted from the functional DWI images can serve as a better alternative to biopsy.

Radiomic analysis on magnetic resonance diffusion weighted image in distinguishing triple-negative breast cancer from other subtypes: a feasibility study

PURPOSE

This work aimed to explore whether radiomic features on magnetic resonance diffusion weighted image (MR DWI) can be used to identify triple-negative breast cancer (TNBC) and other subtypes (non-TNBC).

MATERIALS AND METHODS

This retrospective study included 221 unilateral patients who underwent breast MR imaging prior to neoadjuvant chemotherapy. The subtypes of breast cancer include luminal A (n = 63), luminal B (n = 103), human epidermal growth factor receptor-2 (HER2) overexpressing (n = 30), and triple negative (n = 25). Radiomic features were extracted using Omini-Kinetic software on DWI. Student's t-test and Mann-Whitney U test were used to compare the features between TNBC and non-TNBC patients. Logistic regression analysis and receiver operating characteristic (ROC) curve were used to evaluate the diagnostic efficiency of radiomic features. The Fisher discriminant model was employed to distinguish TNBC and non-TNBC patients automatically. An additional validation dataset with 169 patients was utilized to validate the model.

RESULTS

A total of 76 imaging features were extracted from each lesion on DWI images, and 12 radiomic features were statistically significant between TNBC and non-TNBC patients (P < 0.05). The area of receiver operating characteristic curve (AUC) was 0.817 to apply logistic regression analysis. The accuracy of Fisher discriminant model in distinguishing TNBC and non-TNBC patients was 95.4%, and leave-one-out cross validation was achieved with an accuracy of 83.7%. The same classification analysis of the validation dataset showed an accuracy of 83.4% and an AUC of 0.804.

CONCLUSION

Breast lesions exhibit differences in radiomic features from DWI, enabling good discrimination between TNBC and non-TNBC tumors.

Cholangiocarcinoma Evaluation via Imaging and Artificial Intelligence

BACKGROUND

Cholangiocarcinoma (CCA) is a relatively rare malignant biliary system tumor, and yet it represents the second most common primary hepatic neoplasm, following hepatocellular carcinoma. Regardless of the type, location, or etiology, the survival prognosis of these tumors remains poor. The only method of cure for CCA is complete surgical resection, but part of patients with complete resection are still subject to local recurrence or distant metastasis.

SUMMARY

Over the last several decades, our understanding of the molecular biology of CCA has increased tremendously, diagnostic and evaluative techniques have evolved, and novel therapeutic approaches have been established. Key Messages: This review provides an overview of preoperative imaging evaluations of CCA. Furthermore, relevant information about artificial intelligence (AI) in medical imaging is discussed, as well as the development of AI in CCA treatment.

PET Parametric Imaging: Past, Present, and Future

Positron emission tomography (PET) is actively used in a diverse range of applications in oncology, cardiology, and neurology. The use of PET in the clinical setting focuses on static (single time frame) imaging at a specific time-point post radiotracer injection and is typically considered as semi-quantitative; e.g. standardized uptake value (SUV) measures. In contrast, dynamic PET imaging requires increased acquisition times but has the advantage that it measures the full spatiotemporal distribution of a radiotracer and, in combination with tracer kinetic modeling, enables the generation of multiparametric images that more directly quantify underlying biological parameters of interest, such as blood flow, glucose metabolism, and receptor binding. Parametric images have the potential for improved detection and for more accurate and earlier therapeutic response assessment. Parametric imaging with dynamic PET has witnessed extensive research in the past four decades. In this paper, we provide an overview of past and present activities and discuss emerging opportunities in the field of parametric imaging for the future.

MRI Radiomic Features to Predict IDH1 Mutation Status in Gliomas: A Machine Learning Approach using Gradient Tree Boosting

Patients with gliomas, isocitrate dehydrogenase 1 (IDH1) mutation status have been studied as a prognostic indicator. Recent advances in machine learning (ML) have demonstrated promise in utilizing radiomic features to study disease processes in the brain. We investigate whether ML analysis of multiparametric radiomic features from preoperative Magnetic Resonance Imaging (MRI) can predict IDH1 mutation status in patients with glioma. This retrospective study included patients with glioma with known IDH1 status and preoperative MRI. Radiomic features were extracted from Fluid-Attenuated Inversion Recovery (FLAIR) and Diffused Weighted Imaging (DWI). The dataset was split into training, validation, and testing sets by stratified sampling. Synthetic Minority Oversampling Technique (SMOTE) was applied to the training sets. eXtreme Gradient Boosting (XGBoost) classifiers were trained, and the hyperparameters were tuned. Receiver operating characteristic curve (ROC), accuracy, and f1-scores were collected. A total of 100 patients (age: 55 ± 15, M/F 60/40); with IDH1 mutant (n = 22) and IDH1 wildtype (n = 78) were included. The best performance was seen with a DWI-trained XGBoost model, which achieved ROC with Area Under the Curve (AUC) of 0.97, accuracy of 0.90, and f1-score of 0.75 on the test set. The FLAIR-trained XGBoost model achieved ROC with AUC of 0.95, accuracy of 0.90, f1-score of 0.75 on the test set. A model that was trained on combined FLAIR-DWI radiomic features did not provide incremental accuracy. The results show that a XGBoost classifier using multiparametric radiomic features derived from preoperative MRI can predict IDH1 mutation status with > 90% accuracy.

Integrated Multiparametric Radiomics and Informatics System for Characterizing Breast Tumor Characteristics with the OncotypeDX Gene Assay

Optimal use of multiparametric magnetic resonance imaging (mpMRI) can identify key MRI parameters and provide unique tissue signatures defining phenotypes of breast cancer. We have developed and implemented a new machine-learning informatic system, termed Informatics Radiomics Integration System (IRIS) that integrates clinical variables, derived from imaging and electronic medical health records (EHR) with multiparametric radiomics (mpRad) for identifying potential risk of local or systemic recurrence in breast cancer patients. We tested the model in patients (n = 80) who had Estrogen Receptor positive disease and underwent OncotypeDX gene testing, radiomic analysis, and breast mpMRI. The IRIS method was trained using the mpMRI, clinical, pathologic, and radiomic descriptors for prediction of the OncotypeDX risk score. The trained mpRad IRIS model had a 95% and specificity was 83% with an Area Under the Curve (AUC) of 0.89 for classifying low risk patients from the intermediate and high-risk groups. The lesion size was larger for the high-risk group (2.9 ± 1.7 mm) and lower for both low risk (1.9 ± 1.3 mm) and intermediate risk (1.7 ± 1.4 mm) groups. The lesion apparent diffusion coefficient (ADC) map values for high- and intermediate-risk groups were significantly (p < 0.05) lower than the low-risk group (1.14 vs. 1.49 × 10-3 mm2/s). These initial studies provide deeper insight into the clinical, pathological, quantitative imaging, and radiomic features, and provide the foundation to relate these features to the assessment of treatment response for improved personalized medicine.

Optimal co-clinical radiomics: Sensitivity of radiomic features to tumour volume, image noise and resolution in co-clinical T1-weighted and T2-weighted magnetic resonance imaging

Background

Radiomics analyses has been proposed to interrogate the biology of tumour as well as to predict/assess response to therapy in vivo. The objective of this work was to assess the sensitivity of radiomics features to noise, resolution, and tumour volume in the context of a co-clinical trial.

Methods

Triple negative breast cancer (TNBC) patients were recruited into an ongoing co-clinical imaging trial. Sub-typed matched TNBC patient-derived tumour xenografts (PDX) were generated to investigate optimal co-clinical MR radiomic features. The MR imaging protocol included T1-weighed and T2-weighted imaging. To test the sensitivity of radiomics to resolution, PDX were imaged at three different resolutions. Multiple sets of images with varying signal-to-noise ratio (SNR) were generated, and an image independent patch-based method was implemented to measure the noise levels. Forty-eight radiomic features were extracted from manually segmented 2D and 3D segmented tumours and normal tissues of T1- and T2- weighted co-clinical MR images.

Findings

Sixteen radiomics features were identified as volume dependent and corrected for volume-dependency following normalization. Features from grey-level run-length matrix (GLRLM), grey-level size zone matrix (GLSZM) were identified as most sensitive to noise. Radiomic features Kurtosis and Run-length variance (RLV) from GLSZM were most sensitive to changes in resolution in both T1w and T2w MRI. In general, 3D radiomic features were more robust compared to 2D (single slice) measures, although the former exhibited higher variability between subjects.

Interpretation

Tumour volume, noise characteristics, and image resolution significantly impact radiomic analysis in co-clinical studies.

Radiogenomics of lung cancer

Machine learning (ML) and artificial intelligence (AI) are aiding in improving sensitivity and specificity of diagnostic imaging. The rapid adoption of these advanced ML algorithms is transforming imaging analysis; taking us from noninvasive detection of pathology to noninvasive precise diagnosis of the pathology by identifying whether detected abnormality is a secondary to infection, inflammation and/or neoplasm. This is led to the emergence of “Radiobiogenomics”; referring to the concept of identifying biologic (genomic, proteomic) alterations in the detected lesion. Radiobiogenomic involves image segmentation, feature extraction, and ML model to predict underlying tumor genotype and clinical outcomes. Lung cancer is the most common cause of cancer related death worldwide. There are several histologic subtypes of lung cancer, e.g., small cell lung cancer (SCLC), non-small cell lung cancer (NSCLC) (adenocarcinoma, squamous cell carcinoma). These variable histologic subtypes not only appear different at microscopic level, but these also differ at genetic and transcription level. This intrinsic heterogeneity reveals itself as different morphologic appearances on diagnostic imaging, such as CT, PET/CT and MRI. Traditional evaluation of imaging findings of lung cancer is limited to morphologic characteristics, such as lesion size, margins, density. Radiomics takes image analysis a step further by looking at imaging phenotype with higher order statistics in efforts to quantify intralesional heterogeneity. This heterogeneity, in turn, can be potentially used to extract intralesional genomic and proteomic data. This review aims to highlight novel concepts in ML and AI and their potential applications in identifying radiobiogenomics of lung cancer.

Integrative radiogenomics for virtual biopsy and treatment monitoring in ovarian cancer

BACKGROUND

Ovarian cancer survival rates have not changed in the last 20 years. The majority of cases are High-grade serous ovarian carcinomas (HGSOCs), which are typically diagnosed at an advanced stage with multiple metastatic lesions. Taking biopsies of all sites of disease is infeasible, which challenges the implementation of stratification tools based on molecular profiling.

MAIN BODY

In this review, we describe how these challenges might be overcome by integrating quantitative features extracted from medical imaging with the analysis of paired genomic profiles, a combined approach called radiogenomics, to generate virtual biopsies. Radiomic studies have been used to model different imaging phenotypes, and some radiomic signatures have been associated with paired molecular profiles to monitor spatiotemporal changes in the heterogeneity of tumours. We describe different strategies to integrate radiogenomic information in a global and local manner, the latter by targeted sampling of tumour habitats, defined as regions with distinct radiomic phenotypes.

CONCLUSION

Linking radiomics and biological correlates in a targeted manner could potentially improve the clinical management of ovarian cancer. Radiogenomic signatures could be used to monitor tumours during the course of therapy, offering additional information for clinical decision making. In summary, radiogenomics may pave the way to virtual biopsies and treatment monitoring tools for integrative tumour analysis.

Grayscale Ultrasound Radiomic Features and Shear-Wave Elastography Radiomic Features in Benign and Malignant Breast Masses

PURPOSE

 To identify and compare diagnostic performance of radiomic features between grayscale ultrasound (US) and shear-wave elastography (SWE) in breast masses.

MATERIALS AND METHODS

 We retrospectively collected 328 pathologically confirmed breast masses in 296 women who underwent grayscale US and SWE before biopsy or surgery. A representative SWE image of the mass displayed with a grayscale image in split-screen mode was selected. An ROI was delineated around the mass boundary on the grayscale image and copied and pasted to the SWE image by a dedicated breast radiologist for lesion segmentation. A total of 730 candidate radiomic features including first-order statistics and textural and wavelet features were extracted from each image. LASSO regression was used for data dimension reduction and feature selection. Univariate and multivariate logistic regression was performed to identify independent radiomic features, differentiating between benign and malignant masses with calculation of the AUC.

RESULTS

 Of 328 breast masses, 205 (62.5 %) were benign and 123 (37.5 %) were malignant. Following radiomic feature selection, 22 features from grayscale and 6 features from SWE remained. On univariate analysis, all 6 SWE radiomic features (P < 0.0001) and 21 of 22 grayscale radiomic features (P < 0.03) were significantly different between benign and malignant masses. After multivariate analysis, three grayscale radiomic features and two SWE radiomic features were independently associated with malignant breast masses. The AUC was 0.929 for grayscale US and 0.992 for SWE (P < 0.001).

CONCLUSION

 US radiomic features may have the potential to improve diagnostic performance for breast masses, but further investigation of independent and larger datasets is needed.

Breast MRI texture analysis for prediction of BRCA-associated genetic risk

Background

BRCA1/2 deleterious variants account for most of the hereditary breast and ovarian cancer cases. Prediction models and guidelines for the assessment of genetic risk rely heavily on criteria with high variability such as family cancer history. Here we investigated the efficacy of MRI (magnetic resonance imaging) texture features as a predictor for BRCA mutation status.

Methods

A total of 41 female breast cancer individuals at high genetic risk, sixteen with a BRCA1/2 pathogenic variant and twenty five controls were included. From each MRI 4225 computer-extracted voxels were analyzed. Non-imaging features including clinical, family cancer history variables and triple negative receptor status (TNBC) were complementarily used. Lasso-principal component regression (L-PCR) analysis was implemented to compare the predictive performance, assessed as area under the curve (AUC), when imaging features were used, and lasso logistic regression or conventional logistic regression for the remaining analyses.

Results

Lasso-selected imaging principal components showed the highest predictive value (AUC 0.86), surpassing family cancer history. Clinical variables comprising age at disease onset and bilateral breast cancer yielded a relatively poor AUC (~ 0.56). Combination of imaging with the non-imaging variables led to an improvement of predictive performance in all analyses, with TNBC along with the imaging components yielding the highest AUC (0.94). Replacing family history variables with imaging components yielded an improvement of classification performance of ~ 4%, suggesting that imaging compensates the predictive information arising from family cancer structure.

Conclusions

The L-PCR model uncovered evidence for the utility of MRI texture features in distinguishing between BRCA1/2 positive and negative high-risk breast cancer individuals, which may suggest value to diagnostic routine. Integration of computer-extracted texture analysis from MRI modalities in prediction models and inclusion criteria might play a role in reducing false positives or missed cases especially when established risk variables such as family history are missing.

Parallel comparison and combining effect of radiomic and emerging genomic data for prognostic stratification of non-small cell lung carcinoma patients

BACKGROUND

A single institution retrospective analysis of 124 non-small cell lung carcinoma (NSCLC) patients was performed to identify whether disease-free survival (DFS) achieves incremental values when radiomic and genomic data are combined with clinical information.

METHODS

Using the least absolute shrinkage and selection operator (LASSO) Cox regression method, radiomic and genetic features were reduced in number for selection of the most useful prognostic feature. We created four models using only baseline clinical data, clinical data with selected genetic features, clinical data with selected radiomic features, and clinical data with selected genetic and radiomic features together. Multivariate Cox proportional hazards analysis was performed to determine predictors of DFS. Receiver operating characteristic (ROC) calculation was made to compare the discriminative performance for DFS prediction by four constructed models at the five-year time point.

RESULTS

On precontrast scan, improved discrimination performance was obtained in a merging of selected radiomics and genetics (AUC = 0.8638), compared with clinical data only (AUC = 0.7990), selected genetic features (AUC = 0.8497), and selected radiomic features (AUC = 0.8355). On post-contrast scan, discrimination performance was improved (AUC = 0.8672) compared with the clinical variables (AUC = 0.7913), and selected genetic features (AUC = 0.8376) and selected radiomic features (AUC = 0.8399) were considered.

CONCLUSIONS

The combination of selected radiomic and genomic features improved stratification of NSCLC patients upon survival. Thus, integrating clinicopathologic model with radiomic and genomic features may lead to improved prognostic accuracy compared to conventional clinicopathological data alone.

KEY POINTS

SIGNIFICANT FINDINGS OF THE STUDY: Receiver operating characteristic (ROC) calculation was made to compare the discriminative performance for disease-free survival (DFS). The discriminative performance for DFS was better when combining radiomic and genetic features compared to clinical data only, selected genetic features, and selected radiomic features.

WHAT THIS STUDY ADDS

The combination of selected radiomic and genomic features improved stratification of NSCLC patients upon survival. Thus, integrating a clinicopathological model with radiomic and genomic features may lead to improved prognostic accuracy compared to conventional clinicopathological data alone.

Radiomics for the Diagnosis and Differentiation of Pancreatic Cystic Lesions

Radiomics, also known as quantitative imaging or texture analysis, involves extracting a large number of features traditionally unmeasured in conventional radiological cross-sectional images and converting them into mathematical models. This review describes this approach and its use in the evaluation of pancreatic cystic lesions (PCLs). This discipline has the potential of more accurately assessing, classifying, risk stratifying, and guiding the management of PCLs. Existing studies have provided important insight into the role of radiomics in managing PCLs. Although these studies are limited by the use of retrospective design, single center data, and small sample sizes, radiomic features in combination with clinical data appear to be superior to the current standard of care in differentiating cyst type and in identifying mucinous PCLs with high-grade dysplasia. Combining radiomic features with other novel endoscopic diagnostics, including cyst fluid molecular analysis and confocal endomicroscopy, can potentially optimize the predictive accuracy of these models. There is a need for multicenter prospective studies to elucidate the role of radiomics in the management of PCLs.

Radiomics for Tumor Characterization in Breast Cancer Patients: A Feasibility Study Comparing Contrast-Enhanced Mammography and Magnetic Resonance Imaging

The aim of our intra-individual comparison study was to investigate and compare the potential of radiomics analysis of contrast-enhanced mammography (CEM) and dynamic contrast-enhanced magnetic resonance imaging (DCE-MRI) of the breast for the non-invasive assessment of tumor invasiveness, hormone receptor status, and tumor grade in patients with primary breast cancer. This retrospective study included 48 female patients with 49 biopsy-proven breast cancers who underwent pretreatment breast CEM and MRI. Radiomics analysis was performed by using MaZda software. Radiomics parameters were correlated with tumor histology (invasive vs. non-invasive), hormonal status (HR+ vs. HR-), and grading (low grade G1 + G2 vs. high grade G3). CEM radiomics analysis yielded classification accuracies of up to 92% for invasive vs. non-invasive breast cancers, 95.6% for HR+ vs. HR- breast cancers, and 77.8% for G1 + G2 vs. G3 invasive cancers. MRI radiomics analysis yielded classification accuracies of up to 90% for invasive vs. non-invasive breast cancers, 82.6% for HR+ vs. HR- breast cancers, and 77.8% for G1+G2 vs. G3 cancers. Preliminary results indicate a potential of both radiomics analysis of DCE-MRI and CEM for non-invasive assessment of tumor-invasiveness, hormone receptor status, and tumor grade. CEM may serve as an alternative to MRI if MRI is not available or contraindicated.

Radiomics in radiation oncology-basics, methods, and limitations

Over the past years, the quantity and complexity of imaging data available for the clinical management of patients with solid tumors has increased substantially. Without the support of methods from the field of artificial intelligence (AI) and machine learning, a complete evaluation of the available image information is hardly feasible in clinical routine. Especially in radiotherapy planning, manual detection and segmentation of lesions is laborious, time consuming, and shows significant variability among observers. Here, AI already offers techniques to support radiation oncologists, whereby ultimately, the productivity and the quality are increased, potentially leading to an improved patient outcome. Besides detection and segmentation of lesions, AI allows the extraction of a vast number of quantitative imaging features from structural or functional imaging data that are typically not accessible by means of human perception. These features can be used alone or in combination with other clinical parameters to generate mathematical models that allow, for example, prediction of the response to radiotherapy. Within the large field of AI, radiomics is the subdiscipline that deals with the extraction of quantitative image features as well as the generation of predictive or prognostic mathematical models. This review gives an overview of the basics, methods, and limitations of radiomics, with a focus on patients with brain tumors treated by radiation therapy.

Radiomics in neuro-oncology: Basics, workflow, and applications

Over the last years, the amount, variety, and complexity of neuroimaging data acquired in patients with brain tumors for routine clinical purposes and the resulting number of imaging parameters have substantially increased. Consequently, a timely and cost-effective evaluation of imaging data is hardly feasible without the support of methods from the field of artificial intelligence (AI). AI can facilitate and shorten various time-consuming steps in the image processing workflow, e.g., tumor segmentation, thereby optimizing productivity. Besides, the automated and computer-based analysis of imaging data may help to increase data comparability as it is independent of the experience level of the evaluating clinician. Importantly, AI offers the potential to extract new features from the routinely acquired neuroimages of brain tumor patients. In combination with patient data such as survival, molecular markers, or genomics, mathematical models can be generated that allow, for example, the prediction of treatment response or prognosis, as well as the noninvasive assessment of molecular markers. The subdiscipline of AI dealing with the computation, identification, and extraction of image features, as well as the generation of prognostic or predictive mathematical models, is termed radiomics. This review article summarizes the basics, the current workflow, and methods used in radiomics with a focus on feature-based radiomics in neuro-oncology and provides selected examples of its clinical application.

Radiomic mapping model for prediction of Ki-67 expression in adrenocortical carcinoma

AIM

To determine the value of contrast-enhanced computed tomography (CT)-derived radiomic features in the preoperative prediction of Ki-67 expression in adrenocortical carcinoma (ACC) and to detect significant associations between radiomic features and Ki-67 expression in ACC.

MATERIALS AND METHODS

For this retrospective analysis, patients with histopathologically proven ACC were reviewed. Radiomic features were extracted for all patients from the preoperative contrast-enhanced abdominal CT images. Statistical analysis identified the radiomic features predicting the Ki-67 index in ACC and analysed the correlation with the Ki-67 index.

RESULTS

Fifty-three cases of ACC that met eligibility criteria were identified and analysed. Of the radiomic features analysed, 10 showed statistically significant differences between the high and low Ki-67 expression subgroups. Multivariate linear regression analysis yielded a predictive model showing a significant association between radiomic signature and Ki-67 expression status in ACC (R²=0.67, adjusted R²=0.462, p=0.002). Further analysis of the independent predictors showed statistically significant correlation between Ki-67 expression and shape flatness, elongation, and grey-level long run emphasis (p=0.002, 0.01, and 0.04, respectively). The area under the curve for identification of high Ki-67 expression status was 0.78 for shape flatness and 0.7 for shape elongation.

CONCLUSION

Radiomic features derived from preoperative contrast-enhanced CT images show encouraging results in the prediction of the Ki-67 index in patients with ACC. Morphological features, such as shape flatness and elongation, were superior to other radiomic features in the detection of high Ki-67 expression.

MRI Radiomics for Prediction of Tumor Response and Downstaging in Rectal Cancer Patients after Preoperative Chemoradiation

Purpose

This study aimed to investigate radiomic features extracted from magnetic resonance imaging (MRI) scans performed before and after neoadjuvant chemoradiotherapy (nCRT) in predicting response of locally advanced rectal cancer (LARC).

Methods and Materials

Thirty-nine patients who underwent nCRT for LARC were included, with 294 radiomic features extracted from MRI that was performed before (pre-CRT) and 6 to 8 weeks after completing nCRT (post-CRT). Based on tumor regression grade (TRG), 26 patients were classified as having a histopathologic good response (GR; TRG 0-1) and 13 as non-GR (TRG 2-3). Tumor downstaging (T-downstaging) occurred in 25 patients. Univariate analyses were performed to assess potential radiomic and delta-radiomic predictors for TRG in pathologic complete response (pCR) versus non-pCR, GR versus non-GR, and T-downstaging. The support vector machine-based multivariate model was used to select the best predictors for TRG and T-downstaging.

Results

We identified 13 predictive features for pCR versus non-pCR, 14 for GR versus non-GR, and 16 for T-downstaging. Pre-CRT gray-level run length matrix nonuniformity, pre-CRT neighborhood intensity difference matrix (NIDM) texture strength, and post-CRT NIDM busyness predicted all 3 treatment responses. The best predictor for GR versus non-GR was pre-CRT global minimum combined with clinical N stage in the multivariate analysis. The best predictor for T-downstaging was the combination of pre-CRT gray-level co-occurrence matrix correlation, NIDM-texture strength, and gray-level co-occurrence matrix variance. The pre-CRT, post-CRT, and delta radiomic-based models had no significant difference in predicting all 3 responses.

Conclusions

Pre-CRT MRI, post-CRT MRI, and delta radiomic-based models have the potential to predict tumor response after nCRT in LARC. These data, if validated in larger cohorts, can provide important predictive information to aid in clinical decision making.

Electron Density and Biologically Effective Dose (BED) Radiomics-Based Machine Learning Models to Predict Late Radiation-Induced Subcutaneous Fibrosis

Purpose: to predict the occurrence of late subcutaneous radiation induced fibrosis (RIF) after partial breast irradiation (PBI) for breast carcinoma by using machine learning (ML) models and radiomic features from 3D Biologically Effective Dose (3D-BED) and Relative Electron Density (3D-RED). Methods: 165 patients underwent external PBI following a hypo-fractionation protocol consisting of 40 Gy/10 fractions, 35 Gy/7 fractions, and 28 Gy/4 fractions, for 73, 60, and 32 patients, respectively. Physicians evaluated toxicity at regular intervals by the Common Terminology Adverse Events (CTAE) version 4.0. RIF was assessed every 3 months after the completion of radiation course and scored prospectively. RIF was experienced by 41 (24.8%) patients after average 5 years of follow up. The Hounsfield Units (HU) of the CT-images were converted into relative electron density (3D-RED) and Dose maps into Biologically Effective Dose (3D-BED), respectively. Shape, first-order and textural features of 3D-RED and 3D-BED were calculated in the planning target volume (PTV) and breast. Clinical and demographic variables were also considered (954 features in total). Imbalance of the dataset was addressed by data augmentation using ADASYN technique. A subset of non-redundant features that best predict the data was identified by sequential feature selection. Support Vector Machines (SVM), ensemble machine learning (EML) using various aggregation algorithms and Naive Bayes (NB) classifiers were trained on patient dataset to predict RIF occurrence. Models were assessed using sensitivity and specificity of the ML classifiers and the area under the receiver operator characteristic curve (AUC) of the score functions in repeated 5-fold cross validation on the augmented dataset. Results: The SVM model with seven features was preferred for RIF prediction and scored sensitivity 0.83 (95% CI 0.80-0.86), specificity 0.75 (95% CI 0.71-0.77) and AUC of the score function 0.86 (0.85-0.88) on cross-validation. The selected features included cluster shade and Run Length Non-uniformity of breast 3D-BED, kurtosis and cluster shade from PTV 3D-RED, and 10th percentile of PTV 3D-BED. Conclusion: Textures extracted from 3D-BED and 3D-RED in the breast and PTV can predict late RIF and may help better select patient candidates to exclusive PBI.

Bridging the gap between micro- and macro-scales in medical imaging with textural analysis - A biological basis for CT radiomics classifiers?

INTRODUCTION

Studies suggest there is utility in computed tomography (CT) radiomics for pancreatic disease; however, the precise biological interpretation of its features is unclear. In this manuscript, we present a novel approach towards this interpretation by investigating sub-micron tissue structure using digital pathology.

METHODS

A classification-to attenuation (CAT) function was developed and applied to digital pathology images to create sub-micron linear attenuation maps. From these maps, grey level co-occurrence matrix (GLCM) features were extracted and compared to pathology features. To simulate the spatial frequency loss in a CT scanner, the attenuation maps were convolved with a point spread function (PSF) and subsequently down-sampled. GLCM features were extracted from these down-sampled maps to assess feature stability as a function of spatial frequency loss.

RESULTS

Two GLCM features were shown to be strongly and positively correlated (r = 0.8) with underlying characteristics of the tumor microenvironment, namely percent pimonidazole staining in the tumor. All features underwent marked change as a function of spatial frequency loss; progressively larger spatial frequency losses resulted in progressively larger inter-tumor standard deviations; two GLCM features exhibited stability up to a 100 µm pixel size.

CONCLUSION

This work represents a necessary step towards understanding the biological significance of radiomics. Our preliminary results suggest that cellular metrics of pimonidazole-detectable hypoxia correlate with sub-micron attenuation coefficient texture; however, the consistency of these textures in face of spatial frequency loss is detrimental for robust radiomics. Further study in larger data sets may elucidate additional, potentially more robust features of biologic and clinical relevance.

Radiomics and Deep Learning: Hepatic Applications

Radiomics and deep learning have recently gained attention in the imaging assessment of various liver diseases. Recent research has demonstrated the potential utility of radiomics and deep learning in staging liver fibroses, detecting portal hypertension, characterizing focal hepatic lesions, prognosticating malignant hepatic tumors, and segmenting the liver and liver tumors. In this review, we outline the basic technical aspects of radiomics and deep learning and summarize recent investigations of the application of these techniques in liver disease.

An ensemble learning approach for brain cancer detection exploiting radiomic features

BACKGROUND AND OBJECTIVE

The brain cancer is one of the most aggressive tumour: the 70% of the patients diagnosed with this malignant cancer will not survive. Early detection of brain tumours can be fundamental to increase survival rates. The brain cancers are classified into four different grades (i.e., I, II, III and IV) according to how normal or abnormal the brain cells look. The following work aims to recognize the different brain cancer grades by analysing brain magnetic resonance images.

METHODS

A method to identify the components of an ensemble learner is proposed. The ensemble learner is focused on the discrimination between different brain cancer grades using non invasive radiomic features. The considered radiomic features are belonging to five different groups: First Order, Shape, Gray Level Co-occurrence Matrix, Gray Level Run Length Matrix and Gray Level Size Zone Matrix. We evaluate the features effectiveness through hypothesis testing and through decision boundaries, performance analysis and calibration plots thus we select the best candidate classifiers for the ensemble learner.

RESULTS

We evaluate the proposed method with 111,205 brain magnetic resonances belonging to two freely available data-sets for research purposes. The results are encouraging: we obtain an accuracy of 99% for the benign grade I and the II, III and IV malignant brain cancer detection.

CONCLUSION

The experimental results confirm that the ensemble learner designed with the proposed method outperforms the current state-of-the-art approaches in brain cancer grade detection starting from magnetic resonance images.

Multiparametric radiomics methods for breast cancer tissue characterization using radiological imaging

BACKGROUND AND PURPOSE

Multiparametric radiological imaging is vital for detection, characterization, and diagnosis of many different diseases. Radiomics provide quantitative metrics from radiological imaging that may infer potential biological meaning of the underlying tissue. However, current methods are limited to regions of interest extracted from a single imaging parameter or modality, which limits the amount of information available within the data. This limitation can directly affect the integration and applicable scope of radiomics into different clinical settings, since single image radiomics are not capable of capturing the true underlying tissue characteristics in the multiparametric radiological imaging space. To that end, we developed a multiparametric imaging radiomic (mpRad) framework for extraction of first and second order radiomic features from multiparametric radiological datasets.

METHODS

We developed five different radiomic techniques that extract different aspects of the inter-voxel and inter-parametric relationships within the high-dimensional multiparametric magnetic resonance imaging breast datasets. Our patient cohort consisted of 138 breast patients, where, 97 patients had malignant lesions and 41 patients had benign lesions. Sensitivity, specificity, receiver operating characteristic (ROC) and areas under the curve (AUC) analysis were performed to assess diagnostic performance of the mpRad parameters. Statistical significance was set at p < 0.05.

RESULTS

The mpRad features successfully classified malignant from benign breast lesions with excellent sensitivity and specificity of 82.5% and 80.5%, respectively, with Area Under the receiver operating characteristic Curve (AUC) of 0.87 (0.81-0.93). mpRad provided a 9-28% increase in AUC metrics over single radiomic parameters.

CONCLUSIONS

We have introduced the mpRad framework that extends radiomic analysis from single images to multiparametric datasets for better characterization of the underlying tissue biology.

Radiomics of peripheral nerves MRI in mild carpal and cubital tunnel syndrome

OBJECTIVE

To assess the discriminative power of radiomics of peripheral nerves at 1.5T MRI, using common entrapment neuropathies of the upper limb as a model system of focal nerve injury.

MATERIALS AND METHODS

Radiomics was retrospectively done on peripheral nerve fascicles on T1-weighted 1.5T MRI of 40 patients with diagnosis of mild carpal (n = 25) and cubital tunnel (n = 15) syndrome and of 200 controls. Z-score normalization and Mann-Whitney U test were used to compare features of normal and pathological peripheral nerves. Receiver operating characteristic analysis was performed.

RESULTS

A total of n = 104 radiomics features were computed for each patient and control. Significant differences between normal and pathological median and ulnar nerves were found in n = 23/104 features (p < 0.001). According to features classification, n = 5/23 features were shape-based, n = 7/23 were first-order features, n = 11/23 features were classified as gray level run length matrix. Nine of the selected features showed an AUC higher that 0.7: minimum AUC of 0.74 (95% CI 0.61-0.89) for sum variance and maximum AUC of 0.90 (95% CI 0.82-0.99) for zone entropy.

CONCLUSION

Features analysis demonstrated statistically significant differences between normal and pathological nerve. The results suggested that radiomics analysis could assess the median and ulnar nerve inner structure changes due to the loss of the fascicular pattern, intraneural edema, fibrosis or fascicular alterations in mild carpal tunnel and mild cubital tunnel syndromes even when the nerve cross-sectional area does not change.

Precise hepatectomy in the intelligent digital era

In the past 20 years, the concept of surgery has undergone profound changes. Surgical practice has shifted from emphasizing the complete elimination of lesions to achieving optimal rehabilitation in patients. Collaborative optimization of surgery consists of three core elements, removal of lesions, organ protection and injury close monitoring, and controlled surgical intervention. As a result, the traditional surgical paradigm has quietly transformed into a modern precision surgical paradigm. In this review, we summarized the latest breakthroughs and applications of precision medicine in liver surgery. In addition, we also outlined the progresses that have been made in precision liver surgery, the opportunities and challenges that may encountered in the future.

Initial assessment of 3D magnetic resonance fingerprinting (MRF) towards quantitative brain imaging for radiation therapy

PURPOSE

Magnetic resonance fingerprinting (MRF) provides quantitative T1/T2 maps, enabling applications in clinical radiotherapy such as large-scale, multi-center clinical trials for longitudinal assessment of therapy response. We evaluated the feasibility of a quantitative three-dimensional-MRF (3D-MRF) towards its radiotherapy applications of primary brain tumors.

METHODS

A fast whole-brain 3D-MRF sequence initially developed for diagnostic radiology was optimized using flexible body coils, which is the typical MR imaging setup for radiotherapy treatment planning and for MR imaging (MRI)-guided treatment delivery. Optimization criteria included the accuracy and the precision of T1/T2 quantifications of polyvinylpyrrolidone (PVP) solutions, compared to those from the 3D-MRF using a 32-channel head coil. The accuracy of T1/T2 quantifications from the optimized MRF was first examined in healthy volunteers with two different coil setups. The intra- and inter-scanner variations of image intensity from the optimized sequence were quantified by longitudinal scans of the PVP solutions on two 3T scanners. Using a 3D-printed MRI geometry phantom, susceptibility-induced distortion with the optimized 3D-MRF was quantified as the Dice coefficient of phantom contours, compared to those from CT images. By introducing intentional head motion during 10% of the scan, the robustness of the optimized 3D-MRF towards motion was evaluated through visual inspection of motion artifacts and through quantitative analysis of image sharpness in brain MRF maps.

RESULTS

The optimized sequence acquired whole-brain T1, T2 and proton density maps and with a resolution of 1.2 × 1.2 × 3 mm³ in 10 min, similar to the total acquisition time of 3D T1- and T2-weighted images of the same resolution. In vivo T1 and T2 values of the white and gray matter were consistent with literature. The intra- and inter-scanner variability of the intensity-normalized MRF T1 was 1.0% ± 0.7% and 2.3% ± 1.0% respectively, in contrast to 5.3% ± 3.8% and 3.2% ± 1.6% from the normalized T1-weighted MRI. Repeatability and reproducibility of MRF T1 were independent of intensity normalization. Both phantom and human data demonstrated that the optimized 3D-MRF is more robust to subject motion and artifacts from subject-specific susceptibility difference. Compared to CT contours, the Dice coefficient of phantom contours from 3D-MRF was 0.93, improved from 0.87 from the T1-weighted MRI.

CONCLUSION

Compared to conventional MRI, the optimized 3D-MRF demonstrated improved repeatability across time points and reproducibility across scanners for better tissue quantification, as well as improved robustness to subject-specific susceptibility and motion artifacts under a typical MR imaging setup for radiotherapy. More importantly, quantitative MRF T1/T2 measurements lead to promising potentials towards longitudinal quantitative assessment of treatment response for better adaptive therapy and for large-scale, multi-center clinical trials.

Measurement Variability in Treatment Response Determination for Non-Small Cell Lung Cancer: Improvements Using Radiomics

Multimodality imaging measurements of treatment response are critical for clinical practice, oncology trials, and the evaluation of new treatment modalities. The current standard for determining treatment response in non-small cell lung cancer (NSCLC) is based on tumor size using the RECIST criteria. Molecular targeted agents and immunotherapies often cause morphological change without reduction of tumor size. Therefore, it is difficult to evaluate therapeutic response by conventional methods. Radiomics is the study of cancer imaging features that are extracted using machine learning and other semantic features. This method can provide comprehensive information on tumor phenotypes and can be used to assess therapeutic response in this new age of immunotherapy. Delta radiomics, which evaluates the longitudinal changes in radiomics features, shows potential in gauging treatment response in NSCLC. It is well known that quantitative measurement methods may be subject to substantial variability due to differences in technical factors and require standardization. In this review, we describe measurement variability in the evaluation of NSCLC and the emerging role of radiomics.

A New Application of Multimodality Radiomics Improves Diagnostic Accuracy of Nonpalpable Breast Lesions in Patients with Microcalcifications-Only in Mammography

Background

The aim of this study was to assess a radiomic scheme that combines image features from digital mammography and dynamic contrast-enhanced MRI to improve classification accuracy of nonpalpable breast lesion (NBL) with Breast Imaging-Reporting and Data System (BI-RADS) 3–5 microcalcifications-only in mammography.

Material/Methods

This retrospective study was approved by the Internal Research Review and Ethical Committee of our hospital. We included 81 patients who underwent a three-dimensional digital breast X-ray wire positioning for local resection between October 2012 and November 2016. All patients underwent breast MRI and mammography before the treatment, and all obtained pathological confirmation. According to the pathological results, 41 patients with benign lesions were assigned to the benign group and 40 patients with malignant lesions were assigned to the malignant group. We used the random forest algorithm to select significant features and to test the single and multimodal classifiers using the Leave-One-Out-Cross-Validation method. An area under the receiver operating characteristic curve was also used to evaluate its discriminating performance.

Results

The multimodal classifier achieved AUC of 0.903, with a sensitivity of 82.5% and a specificity of 80.48%, which was better than any single modality.

Conclusions

Multimodal radiomics classification shows promising power in discriminating malignant lesions from benign lesions in NBL patients with BI-RADS 3–5 microcalcifications-only in mammography.