Sonoelastomics for Breast Tumor Classification: A Radiomics Approach with Clustering-Based Feature Selection on Sonoelastography

A combined model based on radiomics features of Sonazoid contrast-enhanced ultrasound in the Kupffer phase for the diagnosis of well-differentiated hepatocellular carcinoma and atypical focal liver lesions: a prospective, multicenter study

OBJECTIVE

The objective of this study was to develop a combined model based on radiomics features of Sonazoid contrast-enhanced ultrasound (CEUS) during the Kupffer phase and to evaluate its value in differentiating well-differentiated hepatocellular carcinoma (w-HCC) from atypical benign focal liver lesions (FLLs).

METHODS

A total of 116 patients with preoperatively Sonazoid-CEUS confirmed w-HCC or benign FLL were selected from a prospective multiple study on the clinical application of Sonazoid in FLLs conducted from August 2020 to March 2021. According to the randomization principle, the patients were divided into a training cohort and a test cohort in a 7:3 ratio. Seventy-nine patients were used for establishing and training the radiomics model and combined model. In comparison, 37 patients were used for validating and comparing the performance of the models. The diagnostic efficacy of the models for w-HCC and atypical benign FLLs was evaluated using ROCs curves and decision curves. A combined model nomogram was created to assess its value in reducing unnecessary biopsies.

RESULTS

Among the patients, there were 55 cases of w-HCC and 61 cases of atypical benign FLLs, including 28 cases of early liver abscess, 16 cases of atypical hepatic hemangioma, 8 cases of hepatocellular dysplastic nodules (DN), and 9 cases of focal nodular hyperplasia (FNH). The radiomics model and combined model we established had AUCs of 0.905 and 0.951, respectively, in the training cohort, and the AUCs of the two models in the test cohort were 0.826 and 0.912, respectively. The combined model outperformed the radiomics feature model significantly. Decision curve analysis demonstrated that the combined model achieved a higher net benefit within a specific threshold probability range (0.25 to 1.00). A nomogram of the combined model was developed.

CONCLUSION

The combined model based on the radiomics features of Sonazoid-CEUS in the Kupffer phase showed satisfactory performance in diagnosing w-HCC and atypical benign FLLs. It can assist clinicians in timely detecting malignant FLLs and reducing unnecessary biopsies for benign diseases.

An Interpretable Radiomics Model Based on Two-Dimensional Shear Wave Elastography for Predicting Symptomatic Post-Hepatectomy Liver Failure in Patients with Hepatocellular Carcinoma

OBJECTIVE

The aim of this study was to develop and validate an interpretable radiomics model based on two-dimensional shear wave elastography (2D-SWE) for symptomatic post-hepatectomy liver failure (PHLF) prediction in patients undergoing liver resection for hepatocellular carcinoma (HCC).

METHODS

A total of 345 consecutive patients were enrolled. A five-fold cross-validation was performed during training, and the models were evaluated in the independent test cohort. A multi-patch radiomics model was established based on the 2D-SWE images for predicting symptomatic PHLF. Clinical features were incorporated into the models to train the clinical-radiomics model. The radiomics model and the clinical-radiomics model were compared with the clinical model comprising clinical variables and other clinical predictive indices, including the model for end-stage liver disease (MELD) score and albumin-bilirubin (ALBI) score. Shapley Additive exPlanations (SHAP) was used for post hoc interpretability of the radiomics model.

RESULTS

The clinical-radiomics model achieved an AUC of 0.867 (95% CI 0.787-0.947) in the five-fold cross-validation, and this score was higher than that of the clinical model (AUC: 0.809; 95% CI: 0.715-0.902) and the radiomics model (AUC: 0.746; 95% CI: 0.681-0.811). The clinical-radiomics model showed an AUC of 0.822 in the test cohort, higher than that of the clinical model (AUC: 0.684, p = 0.007), radiomics model (AUC: 0.784, p = 0.415), MELD score (AUC: 0.529, p < 0.001), and ALBI score (AUC: 0.644, p = 0.016). The SHAP analysis showed that the first-order radiomics features, including first-order maximum 64 × 64, first-order 90th percentile 64 × 64, and first-order 10th percentile 32 × 32, were the most important features for PHLF prediction.

CONCLUSION

An interpretable clinical-radiomics model based on 2D-SWE and clinical variables can help in predicting symptomatic PHLF in HCC.

Radiomics to Differentiate Malignant and Benign Breast Lesions: A Systematic Review and Diagnostic Test Accuracy Meta-Analysis

Breast cancer is a prevalent global health concern, necessitating accurate diagnostic tools for effective management. Diagnostic imaging plays a pivotal role in breast cancer diagnosis, staging, treatment planning, and outcome evaluation. Radiomics is an emerging field of study in medical imaging that contains a broad set of computational methods to extract quantitative features from radiographic images. This can be utilized to guide diagnosis, treatment response, and prognosis in clinical settings.  A systematic review was performed in concordance with Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) guidelines and the Cochrane Handbook for Systematic Reviews of Diagnostic Test Accuracy. Quality was assessed using the radiomics quality score. Diagnostic sensitivity and specificity of radiomics analysis, with 95% confidence intervals (CIs), were included for meta-analysis. The area under the curve analysis was recorded. An extensive statistical analysis was performed following the Cochrane guidelines. Statistical significance was determined if p-values were less than 0.05. Statistical analyses were conducted using Review Manager (RevMan), Version 5.4.1. A total of 31 manuscripts involving 8,773 patients were included, with 17 contributing to the meta-analysis. The cohort comprised 56.2% malignant breast cancers and 43.8% benign breast lesions. MRI demonstrated a sensitivity of 0.91 (95% CI: 0.89-0.92) and a specificity of 0.84 (95% CI: 0.82-0.86) in differentiating between benign and malignant breast cancers. Mammography-based radiomic features predicted breast cancer subtype with a sensitivity of 0.79 (95% CI: 0.76-0.82) and a specificity of 0.81 (95% CI: 0.79-0.84). Ultrasound-based analysis yielded a sensitivity of 0.92 (95% CI: 0.90-0.94) and a specificity of 0.85 (95% CI: 0.83-0.88). Only one study reported the results of radiomic evaluation from CT, which had a sensitivity of 0.95 (95% CI: 0.88-0.99) and a specificity of 0.56 (95% CI: 0.45-0.67).  Across different imaging modalities, radiomics exhibited robust diagnostic accuracy in differentiating benign and malignant breast lesions. The results underscore the potential of radiomic assessment as a minimally invasive alternative or adjunctive diagnostic tool for breast cancer. This is pioneering data that reports on a novel diagnostic approach that is understudied and underreported. However, due to study limitations, the complexity of this technology, and the need for future development, biopsy still remains the current gold standard method of determining breast cancer type.

A Radiomics Study: Classification of Breast Lesions by Textural Features from Mammography Images

This study investigates the feasibility of using texture radiomics features extracted from mammography images to distinguish between benign and malignant breast lesions and to classify benign lesions into different categories and determine the best machine learning (ML) model to perform the tasks. Six hundred and twenty-two breast lesions from 200 retrospective patient data were segmented and analysed. Three hundred fifty radiomics features were extracted using the Standardized Environment for Radiomics Analysis (SERA) library, one of the radiomics implementations endorsed by the Image Biomarker Standardisation Initiative (IBSI). The radiomics features and selected patient characteristics were used to train selected machine learning models to classify the breast lesions. A fivefold cross-validation was used to evaluate the performance of the ML models and the top 10 most important features were identified. The random forest (RF) ensemble gave the highest accuracy (89.3%) and positive predictive value (66%) and likelihood ratio of 13.5 in categorising benign and malignant lesions. For the classification of benign lesions, the RF model again gave the highest likelihood ratio of 3.4 compared to the other models. Morphological and textural radiomics features were identified as the top 10 most important features from the random forest models. Patient age was also identified as one of the significant features in the RF model. We concluded that machine learning models trained against texture-based radiomics features and patient features give reasonable performance in differentiating benign versus malignant breast lesions. Our study also demonstrated that the radiomics-based machine learning models were able to emulate the visual assessment of mammography lesions, typically used by radiologists, leading to a better understanding of how the machine learning model arrive at their decision.

Artificial Intelligence-Assisted Processing of Anterior Segment OCT Images in the Diagnosis of Vitreoretinal Lymphoma

Anterior segment optical coherence tomography (AS-OCT) allows the explore not only the anterior chamber but also the front part of the vitreous cavity. Our cross-sectional single-centre study investigated whether AS-OCT can distinguish between vitreous involvement due to vitreoretinal lymphoma (VRL) and vitritis in uveitis. We studied AS-OCT images from 28 patients (11 with biopsy-proven VRL and 17 with differential diagnosis uveitis) using publicly available radiomics software written in MATLAB. Patients were divided into two balanced groups: training and testing. Overall, 3260/3705 (88%) AS-OCT images met our defined quality criteria, making them eligible for analysis. We studied five different sets of grey-level samplings (16, 32, 64, 128, and 256 levels), finding that 128 grey levels performed the best. We selected the five most effective radiomic features ranked by the ability to predict the class (VRL or uveitis). We built a classification model using the xgboost python function; through our model, 87% of eyes were correctly diagnosed as VRL or uveitis, regardless of exam technique or lens status. Areas under the receiver operating characteristic curves (AUC) in the 128 grey-level model were 0.95 [CI 0.94, 0.96] and 0.84 for training and testing datasets, respectively. This preliminary retrospective study highlights how AS-OCT can support ophthalmologists when there is clinical suspicion of VRL.

An integrated radiomics nomogram based on conventional ultrasound improves discriminability between fibroadenoma and pure mucinous carcinoma in breast

Objective

To evaluate the ability of integrated radiomics nomogram based on ultrasound images to distinguish between breast fibroadenoma (FA) and pure mucinous carcinoma (P-MC).

Methods

One hundred seventy patients with FA or P-MC (120 in the training set and 50 in the test set) with definite pathological confirmation were retrospectively enrolled. Four hundred sixty-four radiomics features were extracted from conventional ultrasound (CUS) images, and radiomics score (Radscore) was constructed using the Least Absolute Shrinkage and Selection Operator (LASSO) algorithm. Different models were developed by a support vector machine (SVM), and the diagnostic performance of the different models was assessed and validated. A comparison of the receiver operating characteristic (ROC) curve, calibration curve, and decision curve analysis (DCA) was performed to evaluate the incremental value of the different models.

Results

Finally, 11 radiomics features were selected, and then Radscore was developed based on them, which was higher in P-MC in both cohorts. In the test group, the clinic + CUS + radiomics (Clin + CUS + Radscore) model achieved a significantly higher area under the curve (AUC) value (AUC = 0.86, 95% CI, 0.733-0.942) when compared with the clinic + radiomics (Clin + Radscore) (AUC = 0.76, 95% CI, 0.618-0.869, P > 0.05), clinic + CUS (Clin + CUS) (AUC = 0.76, 95% CI, 0.618-0.869, P< 0.05), Clin (AUC = 0.74, 95% CI, 0.600-0.854, P< 0.05), and Radscore (AUC = 0.64, 95% CI, 0.492-0.771, P< 0.05) models, respectively. The calibration curve and DCA also suggested excellent clinical value of the combined nomogram.

Conclusion

The combined Clin + CUS + Radscore model may help improve the differentiation of FA from P-MC.

Artificial intelligence - based ultrasound elastography for disease evaluation - a narrative review

Ultrasound elastography (USE) provides complementary information of tissue stiffness and elasticity to conventional ultrasound imaging. It is noninvasive and free of radiation, and has become a valuable tool to improve diagnostic performance with conventional ultrasound imaging. However, the diagnostic accuracy will be reduced due to high operator-dependence and intra- and inter-observer variability in visual observations of radiologists. Artificial intelligence (AI) has great potential to perform automatic medical image analysis tasks to provide a more objective, accurate and intelligent diagnosis. More recently, the enhanced diagnostic performance of AI applied to USE have been demonstrated for various disease evaluations. This review provides an overview of the basic concepts of USE and AI techniques for clinical radiologists and then introduces the applications of AI in USE imaging that focus on the following anatomical sites: liver, breast, thyroid and other organs for lesion detection and segmentation, machine learning (ML) - assisted classification and prognosis prediction. In addition, the existing challenges and future trends of AI in USE are also discussed.

B-mode ultrasound based CAD for liver cancers via multi-view privileged information learning

B-mode ultrasound-based computer-aided diagnosis model can help sonologists improve the diagnostic performance for liver cancers, but it generally suffers from the bottleneck due to the limited structure and internal echogenicity information in B-mode ultrasound images. Contrast-enhanced ultrasound images provide additional diagnostic information on dynamic blood perfusion of liver lesions for B-mode ultrasound images with improved diagnostic accuracy. Since transfer learning has indicated its effectiveness in promoting the performance of target computer-aided diagnosis model by transferring knowledge from related imaging modalities, a multi-view privileged information learning framework is proposed to improve the diagnostic accuracy of the single-modal B-mode ultrasound-based diagnosis for liver cancers. This framework can make full use of the shared label information between the paired B-mode ultrasound images and contrast-enhanced ultrasound images to guide knowledge transfer It consists of a novel supervised dual-view deep Boltzmann machine and a new deep multi-view SVM algorithm. The former is developed to implement knowledge transfer from the multi-phase contrast-enhanced ultrasound images to the B-mode ultrasound-based diagnosis model via a feature-level learning using privileged information paradigm, which is totally different from the existing learning using privileged information paradigm that performs knowledge transfer in the classifier. The latter further fuses and enhances feature representation learned from three pre-trained supervised dual-view deep Boltzmann machine networks for the classification task. An experiment is conducted on a bimodal ultrasound liver cancer dataset. The experimental results show that the proposed framework outperforms all the compared algorithms with the best classification accuracy of 88.91 ± 1.52%, sensitivity of 88.31 ± 2.02%, and specificity of 89.50 ± 3.12%. It suggests the effectiveness of our proposed MPIL framework for the BUS-based CAD of liver cancers.

Attention-based deep learning for breast lesions classification on contrast enhanced spectral mammography: a multicentre study

BACKGROUND

This study aims to develop an attention-based deep learning model for distinguishing benign from malignant breast lesions on CESM.

METHODS

Preoperative CESM images of 1239 patients, which were definitely diagnosed on pathology in a multicentre cohort, were divided into training and validation sets, internal and external test sets. The regions of interest of the breast lesions were outlined manually by a senior radiologist. We adopted three conventional convolutional neural networks (CNNs), namely, DenseNet 121, Xception, and ResNet 50, as the backbone architectures and incorporated the convolutional block attention module (CBAM) into them for classification. The performance of the models was analysed in terms of the receiver operating characteristic (ROC) curve, accuracy, the positive predictive value (PPV), the negative predictive value (NPV), the F1 score, the precision recall curve (PRC), and heat maps. The final models were compared with the diagnostic performance of conventional CNNs, radiomics models, and two radiologists with specialised breast imaging experience.

RESULTS

The best-performing deep learning model, that is, the CBAM-based Xception, achieved an area under the ROC curve (AUC) of 0.970, a sensitivity of 0.848, a specificity of 1.000, and an accuracy of 0.891 on the external test set, which was higher than those of other CNNs, radiomics models, and radiologists. The PRC and the heat maps also indicated the favourable predictive performance of the attention-based CNN model. The diagnostic performance of two radiologists improved with deep learning assistance.

CONCLUSIONS

Using an attention-based deep learning model based on CESM images can help to distinguishing benign from malignant breast lesions, and the diagnostic performance of radiologists improved with deep learning assistance.

Multi-modality radiomics nomogram based on DCE-MRI and ultrasound images for benign and malignant breast lesion classification

OBJECTIVE

To develop a multi-modality radiomics nomogram based on DCE-MRI, B-mode ultrasound (BMUS) and strain elastography (SE) images for classifying benign and malignant breast lesions.

MATERIAL AND METHODS

In this retrospective study, 345 breast lesions from 305 patients who underwent DCE-MRI, BMUS and SE examinations were randomly divided into training (n = 241) and testing (n = 104) datasets. Radiomics features were extracted from manually contoured images. The inter-class correlation coefficient (ICC), Mann-Whitney U test and the least absolute shrinkage and selection operator (LASSO) regression were applied for feature selection and radiomics signature building. Multivariable logistic regression was used to develop a radiomics nomogram incorporating radiomics signature and clinical factors. The performance of the radiomics nomogram was evaluated by its discrimination, calibration, and clinical usefulness and was compared with BI-RADS classification evaluated by a senior breast radiologist.

RESULTS

The All-Combination radiomics signature derived from the combination of DCE-MRI, BMUS and SE images showed better diagnostic performance than signatures derived from single modality alone, with area under the curves (AUCs) of 0.953 and 0.941 in training and testing datasets, respectively. The multi-modality radiomics nomogram incorporating the All-Combination radiomics signature and age showed excellent discrimination with the highest AUCs of 0.964 and 0.951 in two datasets, respectively, which outperformed all single modality radiomics signatures and BI-RADS classification. Furthermore, the specificity of radiomics nomogram was significantly higher than BI-RADS classification (both p < 0.04) with the same sensitivity in both datasets.

CONCLUSION

The proposed multi-modality radiomics nomogram based on DCE-MRI and ultrasound images has the potential to serve as a non-invasive tool for classifying benign and malignant breast lesions and reduce unnecessary biopsy.

Intelligent diagnosis of left ventricular hypertrophy using transthoracic echocardiography videos

PURPOSE

Left ventricular hypertrophy (LVH) is an independent risk factor for cardiovascular events and mortality. Pathological LVH can be caused by various diseases. In this study, we explored the possibility of using time and frequency domain analysis of myocardial radiomics features for patients with LVH in differentiating hypertrophic cardiomyopathy (HCM), hypertensive heart disease (HHD) and uremic cardiomyopathy (UCM) based on transthoracic echocardiography (TTE). This was the first study to explore TTE myocardial time and frequency domain analyses for multiple LVH etiology differentiation.

MATERIALS AND METHODS

We proposed an artificially intelligent diagnosis system based on radiomics techniques for differentiating HCM, HHD and UCM on TTE videos of the apical four-chamber view, which mainly included interventricular septum (IVS) segmentation, feature extraction and classification. We used two independent cohorts, one with 150 patients, including 50 HHD, 50 HCM and 50 UCM, for segmentation training and testing, and another with 149 patients (namely the main cohort), including 50 HHD, 46 HCM and 53 UCM, for classification training and testing after segmentation and feature extraction. Firstly, the U-Net, Residual U-Net (ResUNet) and nnU-Net were trained and tested to segment the IVS on TTE still images in the first cohort. Then the trained model with the best segmentation performance was further used for IVS prediction of ordered TTE images in video sequences in the main cohort. The post-processing was used to eliminate the noisy debris by selecting the maximum connected region and smoothing the edges of the predicted IVS region. Secondly, static radiomics features were extracted from the IVS of ordered TTE images in each video sequence, and subsequently the time and frequency domain features were further extracted from each time series of a static radiomics feature in the video sequence. Finally, the point-wise gated Boltzmann machine (PGBM) was used to learn and fuse the time and frequency domain features, and the support vector machine was used to classify the learned features for LVH diagnosis. The classification was performed with five-fold cross validation.

RESULTS

The ResUNet showed the best segmentation performance, with Dice coefficient, sensitivity, specificity and accuracy of 0.817, 76.3%, 99.6% and 98.6%, respectively. With post-processing, the Dice coefficient, sensitivity, specificity and accuracy of the ResUNet were further improved to 0.839, 77.0%, 99.8%, and 98.8%, respectively. The classification areas under the receiver operating characteristic curves (AUCs) were 0.838 ± 0.049 for HHD vs. HCM, 0.868 ± 0.042 for HCM vs. UCM and 0.701 ± 0.140 for HHD vs. UCM.

CONCLUSION

In this work, we proposed an intelligent identification system for LVH etiology classification based on routine TTE video images with good diagnostic performance. This deep learning method is feasible in automatic TTE images interpretation and expected to assist clinicians in detecting the primary cause of LVH.

Doubly Supervised Transfer Classifier for Computer-Aided Diagnosis With Imbalanced Modalities

Transfer learning (TL) can effectively improve diagnosis accuracy of single-modal-imaging-based computer-aided diagnosis (CAD) by transferring knowledge from other related imaging modalities, which offers a way to alleviate the small-sample-size problem. However, medical imaging data generally have the following characteristics for the TL-based CAD: 1) The source domain generally has limited data, which increases the difficulty to explore transferable information for the target domain; 2) Samples in both domains often have been labeled for training the CAD model, but the existing TL methods cannot make full use of label information to improve knowledge transfer. In this work, we propose a novel doubly supervised transfer classifier (DSTC) algorithm. In particular, DSTC integrates the support vector machine plus (SVM+) classifier and the low-rank representation (LRR) into a unified framework. The former makes full use of the shared labels to guide the knowledge transfer between the paired data, while the latter adopts the block-diagonal low-rank (BLR) to perform supervised TL between the unpaired data. Furthermore, we introduce the Schatten-p norm for BLR to obtain a tighter approximation to the rank function. The proposed DSTC algorithm is evaluated on the Alzheimer's disease neuroimaging initiative (ADNI) dataset and the bimodal breast ultrasound image (BBUI) dataset. The experimental results verify the effectiveness of the proposed DSTC algorithm.

Diagnostic Strategies for Breast Cancer Detection: From Image Generation to Classification Strategies Using Artificial Intelligence Algorithms

Breast cancer is one the main death causes for women worldwide, as 16% of the diagnosed malignant lesions worldwide are its consequence. In this sense, it is of paramount importance to diagnose these lesions in the earliest stage possible, in order to have the highest chances of survival. While there are several works that present selected topics in this area, none of them present a complete panorama, that is, from the image generation to its interpretation. This work presents a comprehensive state-of-the-art review of the image generation and processing techniques to detect Breast Cancer, where potential candidates for the image generation and processing are presented and discussed. Novel methodologies should consider the adroit integration of artificial intelligence-concepts and the categorical data to generate modern alternatives that can have the accuracy, precision and reliability expected to mitigate the misclassifications.

Prediction of programmed cell death protein 1 in hepatocellular carcinoma patients using radiomics analysis with radiofrequency-based ultrasound multifeature maps

Background

This study explored the feasibility of radiofrequency (RF)-based radiomics analysis techniques for the preoperative prediction of programmed cell death protein 1 (PD-1) in patients with hepatocellular carcinoma (HCC).

Methods

The RF-based radiomics analysis method used ultrasound multifeature maps calculated from the RF signals of HCC patients, including direct energy attenuation (DEA) feature map, skewness of spectrum difference (SSD) feature map, and noncentrality parameter S of the Rician distribution (NRD) feature map. From each of the above ultrasound maps, 345 high-throughput radiomics features were extracted. Then, the useful radiomics features were selected by the sparse representation method and input into support vector machine (SVM) classifier for PD-1 prediction.

Results and conclusion

Among all the RF-based prediction models and the ultrasound grayscale comparative model, the RF-based model using all of the three ultrasound feature maps had the highest prediction accuracy (ACC) and area under the curve (AUC), which were 92.5% and 94.23%, respectively. The method proposed in this paper is effective for the meaningful feature extraction of RF signals and can effectively predict PD-1 in patients with HCC.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12938-021-00927-y.

We proposed RF-based radiomics analysis method by introducing three ultrasound features of direct energy attenuation (DEA), skewness of spectrum difference (SSD) and noncentrality parameter S of Rician distribution (NRD) as the feature extraction method from RF signals, investigated the effectiveness of RF-based radiomics analysis method in the immunocheckpoint prediction of programmed cell death protein 1 (PD-1), and validated the results with contrast testing of grayscale-based radiomics analysis method in this study. We also demonstrate a trend in prediction performance changes and its correlation with the number of ultrasound features.

The results demonstrated that there were significant differences (p < 0.05) in radiomics scores between HCC patients with PD-1 and HCC patients without PD-1. RF-based radiomics analysis method performed well in PD-1 noninvasive preoperative prediction of HCC patients.

In this study, the performance of RF-based radiomics analysis method was better than that of grayscale-based radiomics analysis method in the preoperative prediction of PD-1 in HCC patients. The AUC of DSNM, which was the RF-based radiomics analysis model with three ultrasound feature maps, reached 94.23% in the prediction of PD-1 cell protein in HCC patients.

Exploratory study on the predictive value of ultrasound radiomics for cervical tuberculous lymphadenitis

PURPOSE

To investigate the predictive values of gray-scale ultrasound (G-US) and strain elastic ultrasound (SE-US) radiomic features for cervical tuberculous lymphadenitis (CTL).

MATERIAL AND METHODS

The G-US and SE-US images of 147 patients with pathologically confirmed CTL and 69 non-CTL patients were retrospectively analyzed. A total of 851 imaging features were extracted. The patients were divided into the training set and test set in 7:3 ratio. In the training set, the minimum redundancy maximum relevance (mRMR) and the least absolute shrinkage and selection operator (LASSO) were used for feature selection and modeling. The diagnostic power of G-US and SE-US ultrasound radiomics in identifying CTL was evaluated in test set.

RESULTS

The G-US and SE-US have finally selected 10 and 14 features, respectively. In the G-US group, the diagnostic sensitivity, specificity and accuracy of the training set were 69.7%, 85.7% and 70.0%, respectively, and those values in the test set were 81.3%, 70.0% and 86.4%, respectively. The SE-US group had a sensitivity of 71.7%, a specificity of 81.6%, and an accuracy of 67.0% in the training set, and those parameters in the test set were 81.0%, 75.0%, and 83.7%, respectively. In the G-US group, the positive and negative predictive value of the training set were 0.519 and 0.901, respectively, and those values in the test set were 0.700 and 0.864, respectively. The SE-US group had a positive predictive value of 0.541, and a negative predictive value of 0.885 in the training set, and those parameters in the test set were 0.682 and 0.878, respectively. By Delong test, G-US and SE-US groups showed no significant differences in diagnostic performance between the training and test sets.

CONCLUSIONS

The ultrasound radiomic features of G-US and SE-US exhibited certain predictive potential in detecting CTL, providing a new non-invasive method for clinicians to more accurately evaluate patients with CTL.

Preoperative Prediction Power of Radiomics for Breast Cancer: A Systemic Review and Meta-Analysis

Background

To evaluate the preoperative predictive value of radiomics in the diagnosis of breast cancer (BC).

Methods

By searching PubMed and Embase libraries, our study identified 19 eligible studies. We conducted a meta-analysis to assess the differential value in the preoperative assessment of BC using radiomics methods.

Results

Nineteen radiomics studies focusing on the diagnostic efficacy of BC and involving 5865 patients were enrolled. The integrated sensitivity and specificity were 0.84 (95% CI: 0.80–0.87, I^{2 =} 76.44%) and 0.83 (95% CI: 0.78–0.87, I^{2 =} 81.79%), respectively. The AUC based on the SROC curve was 0.91, indicating a high diagnostic value.

Conclusion

Radiomics has shown excellent diagnostic performance in the preoperative prediction of BC and is expected to be a promising method in clinical practice.

Prediction for pathological and immunohistochemical characteristics of triple-negative invasive breast carcinomas: the performance comparison between quantitative and qualitative sonographic feature analysis

OBJECTIVE

Sonographic features are associated with pathological and immunohistochemical characteristics of triple-negative breast cancer (TNBC). To predict the biological property of TNBC, the performance using quantitative high-throughput sonographic feature analysis was compared with that using qualitative feature assessment.

METHODS

We retrospectively reviewed ultrasound images, clinical, pathological, and immunohistochemical (IHC) data of 252 female TNBC patients. All patients were subgrouped according to the histological grade, Ki67 expression level, and human epidermal growth factor receptor 2 (HER2) score. Qualitative sonographic feature assessment included shape, margin, posterior acoustic pattern, and calcification referring to the Breast Imaging Reporting and Data System (BI-RADS). Quantitative sonographic features were acquired based on the computer-aided radiomics analysis. Breast cancer masses were manually segmented from the surrounding breast tissues. For each ultrasound image, 1688 radiomics features of 7 feature classes were extracted. The principal component analysis (PCA), least absolute shrinkage and selection operator (LASSO), and support vector machine (SVM) were used to determine the high-throughput radiomics features that were highly correlated to biological properties. The performance using both quantitative and qualitative sonographic features to predict biological properties of TNBC was represented by the area under the receiver operating characteristic curve (AUC).

RESULTS

In the qualitative assessment, regular tumor shape, no angular or spiculated margin, posterior acoustic enhancement, and no calcification were used as the independent sonographic features for TNBC. Using the combination of these four features to predict the histological grade, Ki67, HER2, axillary lymph node metastasis (ALNM), and lymphovascular invasion (LVI), the AUC was 0.673, 0.680, 0.651, 0.587, and 0.566, respectively. The number of high-throughput features that closely correlated with biological properties was 34 for histological grade (AUC 0.942), 27 for Ki67 (AUC 0.732), 25 for HER2 (AUC 0.730), 34 for ALNM (AUC 0.804), and 34 for LVI (AUC 0.795).

CONCLUSION

High-throughput quantitative sonographic features are superior to traditional qualitative ultrasound features in predicting the biological behavior of TNBC.

KEY POINTS

• Sonographic appearances of TNBCs showed a great variety in accordance with its biological and clinical characteristics. • Both qualitative and quantitative sonographic features of TNBCs are associated with tumor biological characteristics. • The quantitative high-throughput feature analysis is superior to two-dimensional sonographic feature assessment in predicting tumor biological property.

Machine Learning: Applications and Advanced Progresses of Radiomics in Endocrine Neoplasms

Endocrine neoplasms remain a great threat to human health. It is extremely important to make a clear diagnosis and timely treatment of endocrine tumors. Machine learning includes radiomics, which has long been utilized in clinical cancer research. Radiomics refers to the extraction of valuable information by analyzing a large amount of standard data with high-throughput medical images mainly including computed tomography, positron emission tomography, magnetic resonance imaging, and ultrasound. With the quantitative imaging analysis and model building, radiomics can reflect specific underlying characteristics of a disease that otherwise could not be evaluated visually. More and more promising results of radiomics in oncological practice have been seen in recent years. Radiomics may have the potential to supplement traditional imaging analysis and assist in providing precision medicine for patients. Radiomics had developed rapidly in endocrine neoplasms practice in the past decade. In this review, we would introduce the general workflow of radiomics and summarize the applications and developments of radiomics in endocrine neoplasms in recent years. The limitations of current radiomic research studies and future development directions would also be discussed.

RGB Three-Channel SWE-Based Ultrasomics Model: Improving the Efficiency in Differentiating Focal Liver Lesions

Objective

To explore a new method for color image analysis of ultrasomics and investigate the efficiency in differentiating focal liver lesions (FLLs) by Red, Green, and Blue (RGB) three-channel SWE-based ultrasomics model.

Methods

One hundred thirty FLLs were randomly divided into training set (n = 65) and validation set (n = 65). The RGB three-channel and direct conversion methods were applied to the same color SWE images. Ultrasomics features were extracted from the preprocessing images establishing two feature data sets. The least absolute shrinkage and selection operator (LASSO) logistic regression model was applied for feature selection and model construction. Two models, named RGB model (based on RGB three-channel conversion) and direct model (based on direct conversion), were used to differentiate FLLs. The diagnosis performance of the two models was evaluated by area under the curve (AUC), calibration curves, decision curves, and net reclassification index (NRI).

Results

In the validation cohort, the AUC of the direct model and RGB model in characterization on FLLs were 0.813 and 0.926, respectively (p = 0.038). Calibration curves and decision curves indicated that the RGB model had better calibration efficiency and provided greater clinical benefits. NRI revealed that the RGB model correctly reclassified 7% of malignant cases and 25% of benign cases compared to the direct model (p = 0.01).

Conclusion

The RGB model generated by RGB three-channel method yielded better diagnostic efficiency than the direct model established by direct conversion method. The RGB three-channel method may be promising on ultrasomics analysis of color images in clinical application.

Application of CT-Based Radiomics in Discriminating Pancreatic Cystadenomas From Pancreatic Neuroendocrine Tumors Using Machine Learning Methods

Objectives

The purpose of this study aimed at investigating the reliability of radiomics features extracted from contrast-enhanced CT in differentiating pancreatic cystadenomas from pancreatic neuroendocrine tumors (PNETs) using machine-learning methods.

Methods

In this study, a total number of 120 patients, including 66 pancreatic cystadenomas patients and 54 PNETs patients were enrolled. Forty-eight radiomic features were extracted from contrast-enhanced CT images using LIFEx software. Five feature selection methods were adopted to determine the appropriate features for classifiers. Then, nine machine learning classifiers were employed to build predictive models. The performance of the forty-five models was evaluated with area under the curve (AUC), accuracy, sensitivity, specificity, and F1 score in the testing group.

Results

The predictive models exhibited reliable ability of differentiating pancreatic cystadenomas from PNETs when combined with suitable selection methods. A combination of DC as the selection method and RF as the classifier, as well as Xgboost+RF, demonstrated the best discriminative ability, with the highest AUC of 0.997 in the testing group.

Conclusions

Radiomics-based machine learning methods might be a noninvasive tool to assist in differentiating pancreatic cystadenomas and PNETs.

Oropharyngeal cancer patient stratification using random forest based-learning over high-dimensional radiomic features

To improve risk prediction for oropharyngeal cancer (OPC) patients using cluster analysis on the radiomic features extracted from pre-treatment Computed Tomography (CT) scans. 553 OPC Patients randomly split into training (80%) and validation (20%), were classified into 2 or 3 risk groups by applying hierarchical clustering over the co-occurrence matrix obtained from a random survival forest (RSF) trained over 301 radiomic features. The cluster label was included together with other clinical data to train an ensemble model using five predictive models (Cox, random forest, RSF, logistic regression, and logistic-elastic net). Ensemble performance was evaluated over the independent test set for both recurrence free survival (RFS) and overall survival (OS). The Kaplan-Meier curves for OS stratified by cluster label show significant differences for both training and testing (p val < 0.0001). When compared to the models trained using clinical data only, the inclusion of the cluster label improves AUC test performance from .62 to .79 and from .66 to .80 for OS and RFS, respectively. The extraction of a single feature, namely a cluster label, to represent the high-dimensional radiomic feature space reduces the dimensionality and sparsity of the data. Moreover, inclusion of the cluster label improves model performance compared to clinical data only and offers comparable performance to the models including raw radiomic features.

Diagnosis of Infantile Hip Dysplasia with B-mode Ultrasound via Two-stage Meta-learning Based Deep Exclusivity Regularized Machine

The B-mode ultrasound (BUS) based computer-aided diagnosis (CAD) has shown its effectiveness for developmental dysplasia of the hip (DDH) in infants. In this work, a two-stage meta-learning based deep exclusivity regularized machine (TML-DERM) is proposed for the BUS-based CAD of DDH. TML-DERM integrates deep neural network (DNN) and exclusivity regularized machine into a unified framework to simultaneously improve the feature representation and classification performance. Moreover, the first-stage meta-learning is mainly conducted on the DNN module to alleviate the overfitting issue caused by the significantly increased parameters in DNN, and a random sampling strategy is adopted to self-generate the meta-tasks; while the second-stage meta-learning mainly learns the combination of multiple weak classifiers by a weight vector to improve the classification performance, and also optimizes the unified framework again. The experimental results on a DDH ultrasound dataset show the proposed TML-DERM achieves the superior classification performance with the mean accuracy of 85.89%, sensitivity of 86.54%, and specificity of 85.23%.

Pattern classification for breast lesion on FFDM by integration of radiomics and deep features

The radiomics model can be used in breast cancer detection via calculating quantitative image features. However, these features are explicitly designed, or handcrafted in advance, and this would limit their ability to characterize the lesion properly. This paper aims to build an integrated-features-based classification framework which cooperate the radiomics features and the deep features to classify benign and malignant breast lesions on full-filed digital mammography (FFDM). We propose a classification framework consists of three steps: (1) handcrafted features (HCFs) extraction and selection, (2) deep features (DFs) extraction and (3) the integrated features-based classification. Specifically, HCFs comprise the gray-level gap-length matrix (GLGLM) texture features and shape features, and DFs contain the pooled features and high-level fully-connected features. Then, a multi-classifier method is applied to construct our classification framework using integrated features for breast lesion classification. A total of 106 retrospective FFDM data (51 are malignant and 55 are benign) in both craniocaudal (CC) view and mediolateral oblique (MLO) view were included in this study. The areas under a receiver operating characteristic curve (AUC) value, accuracy, sensitivity, specificity and Youden's index, are used to examine the performance of our proposed method in differentiating benign and malignant breast lesion. Proposed framework trained on the concatenation of fully-connected features and HCFs can significantly improve classification performance (AUC of 94.6 %, accuracy of 96.4 %, sensitivity of 93.6 %, specificity of 98.9 % and Yonden's index of 92.5 %) compared with other features sets. Experimental results demonstrate that performance of proposed framework is improved, indicating the potential of concatenation of the fully-connected features and HCFs set in breast cancer patients.

Images Are Data: A Breast Imaging Perspective on a Contemporary Paradigm

BACKGROUND

 Considering radiological examinations not as mere images, but as a source of data, has become the key paradigm in the diagnostic imaging field. This change of perspective is particularly popular in breast imaging. It allows breast radiologists to apply algorithms derived from computer science, to realize innovative clinical applications, and to refine already established methods. In this context, the terminology "imaging biomarker", "radiomics", and "artificial intelligence" are of pivotal importance. These methods promise noninvasive, low-cost (e. g., in comparison to multigene arrays), and workflow-friendly (automated, only one examination, instantaneous results, etc.) delivery of clinically relevant information.

METHODS AND RESULTS

 This paper is designed as a narrative review on the previously mentioned paradigm. The focus is on key concepts in breast imaging and important buzzwords are explained. For all areas of breast imaging, exemplary studies and potential clinical use cases are discussed.

CONCLUSION

 Considering radiological examination as a source of data may optimize patient management by guiding individualized breast cancer diagnosis and oncologic treatment in the age of precision medicine.

KEY POINTS

  · In conventional breast imaging, examinations are interpreted based on patterns perceivable by visual inspection.. · The radiomics paradigm treats breast images as a source of data, containing information beyond what is visible to our eyes.. · This results in radiomic signatures that may be considered as imaging biomarkers, as they provide diagnostic, predictive, and prognostic information.. · Radiomics derived imaging biomarkers may be used to individualize breast cancer treatment in the era of precision medicine.. · The concept and key research of radiomics in the field of breast imaging will be discussed in this narrative review..

CITATION FORMAT

· Dietzel M, Clauser P, Kapetas P et al. Images Are Data: A Breast Imaging Perspective on a Contemporary Paradigm. Fortschr Röntgenstr 2021; 193: 898 - 908.

A deep learning framework for 18F-FDG PET imaging diagnosis in pediatric patients with temporal lobe epilepsy

PURPOSE

Epilepsy is one of the most disabling neurological disorders, which affects all age groups and often results in severe consequences. Since misdiagnoses are common, many pediatric patients fail to receive the correct treatment. Recently, ¹⁸F-fluorodeoxyglucose positron emission tomography (¹⁸F-FDG PET) imaging has been used for the evaluation of pediatric epilepsy. However, the epileptic focus is very difficult to be identified by visual assessment since it may present either hypo- or hyper-metabolic abnormality with unclear boundary. This study aimed to develop a novel symmetricity-driven deep learning framework of PET imaging for the identification of epileptic foci in pediatric patients with temporal lobe epilepsy (TLE).

METHODS

We retrospectively included 201 pediatric patients with TLE and 24 age-matched controls who underwent ¹⁸F-FDG PET-CT studies. ¹⁸F-FDG PET images were quantitatively investigated using 386 symmetricity features, and a pair-of-cube (PoC)-based Siamese convolutional neural network (CNN) was proposed for precise localization of epileptic focus, and then metabolic abnormality level of the predicted focus was calculated automatically by asymmetric index (AI). Performances of the proposed framework were compared with visual assessment, statistical parametric mapping (SPM) software, and Jensen-Shannon divergence-based logistic regression (JS-LR) analysis.

RESULTS

The proposed deep learning framework could detect the epileptic foci accurately with the dice coefficient of 0.51, which was significantly higher than that of SPM (0.24, P < 0.01) and significantly (or marginally) higher than that of visual assessment (0.31-0.44, P = 0.005-0.27). The area under the curve (AUC) of the PoC classification was higher than that of the JS-LR (0.93 vs. 0.72). The metabolic level detection accuracy of the proposed method was significantly higher than that of visual assessment blinded or unblinded to clinical information (90% vs. 56% or 68%, P < 0.01).

CONCLUSION

The proposed deep learning framework for ¹⁸F-FDG PET imaging could identify epileptic foci accurately and efficiently, which might be applied as a computer-assisted approach for the future diagnosis of epilepsy patients.

TRIAL REGISTRATION

NCT04169581. Registered November 13, 2019 Public site: https://clinicaltrials.gov/ct2/show/NCT04169581.

Radiomics in Breast Imaging from Techniques to Clinical Applications: A Review

Recent advances in computer technology have generated a new area of research known as radiomics. Radiomics is defined as the high throughput extraction and analysis of quantitative features from imaging data. Radiomic features provide information on the gray-scale patterns, inter-pixel relationships, as well as shape and spectral properties of radiological images. Moreover, these features can be used to develop computational models that may serve as a tool for personalized diagnosis and treatment guidance. Although radiomics is becoming popular and widely used in oncology, many problems such as overfitting and reproducibility issues remain unresolved. In this review, we will outline the steps of radiomics used for oncology, specifically addressing applications for breast cancer patients and focusing on technical issues.

Breast cancer, screening and diagnostic tools: All you need to know

Breast cancer is one of the most frequent malignancies among women worldwide. Methods for screening and diagnosis allow health care professionals to provide personalized treatments that improve the outcome and survival. Scientists and physicians are working side-by-side to develop evidence-based guidelines and equipment to detect cancer earlier. However, the lack of comprehensive interdisciplinary information and understanding between biomedical, medical, and technology professionals makes innovation of new screening and diagnosis tools difficult. This critical review gathers, for the first time, information concerning normal breast and cancer biology, established and emerging methods for screening and diagnosis, staging and grading, molecular and genetic biomarkers. Our purpose is to address key interdisciplinary information about these methods for physicians and scientists. Only the multidisciplinary interaction and communication between scientists, health care professionals, technical experts and patients will lead to the development of better detection tools and methods for an improved screening and early diagnosis.

Ultrasound Radiomics Effective for Preoperative Identification of True and Pseudo Gallbladder Polyps Based on Spatial and Morphological Features

Purpose: To explore the value of ultrasound radiomics in the preoperative identification of true and pseudo gallbladder polyps and to evaluate the associated diagnostic accuracy.

Methods: Totally, 99 pathologically proven gallbladder polyps in 96 patients were enrolled, including 58 cholesterol polyps (55 patients) and 41 gallbladder tubular adenomas (41 patients). Features on preoperative ultrasound images, including spatial and morphological features, were acquired for each lesion. Following this, two-stage feature selection was adopted using Fisher's inter-intraclass variance ratios and Z-scores for the selection of intrinsic features important for differential diagnosis achievement with support vector machine use.

Results: Eighty radiomic features were extracted from each polyp. Eight intrinsic features were identified after two-stage selection. The contrast 14 (Cont14) and entropy 6 (Entr6) values in the cholesterol polyp group were significantly higher than those in the gallbladder adenoma group (4.063 ± 1.682 vs. 2.715 ± 1.867, p < 0.001 for Cont14; 4.712 ± 0.427 vs. 4.380 ± 0.720, p = 0.003 for Entr6); however, the homogeneity 13 (Homo13) and energy 8 (Ener8) values in the cholesterol polyp group were significantly lower (0.500 ± 0.069 vs. 0.572 ± 0.057, p < 0.001 for Homo13; 0.050 ± 0.023 vs. 0.068 ± 0.038, p = 0.002 for Ener8). These results indicate that the pixel distribution of cholesterol polyps was more uneven than that of gallbladder tubular adenomas. The dispersion degree was also significantly lower in the cholesterol polyp group than the gallbladder adenoma group (0.579 ± 0.054 vs. 0.608 ± 0.041, p = 0.005), indicating a lower dispersion of high-intensity areas in the cholesterol polyps. The long axis length of the fitting ellipse (Maj.Len), diameter of a circle equal to the lesion area (Eq.Dia) and perimeter (Per) values in the cholesterol polyp group were significantly lower than those in the gallbladder adenoma group (0.971 ± 0.485 vs. 1.738 ± 0.912, p < 0.001 for Maj.Len; 0.818 ± 0.393 vs. 1.438 ± 0.650, p < 0.001 for Eq.Dia; 2.637 ± 1.281 vs. 5.033 ± 2.353, p < 0.001 for Per), demonstrating that the cholesterol polyps were smaller and more regular in terms of morphology. The classification accuracy, sensitivity, specificity, and area under the curve values were 0.875, 0.885, 0.857, and 0.898, respectively.

Conclusions: Ultrasound radiomic analysis based on the spatial and morphological features extracted from ultrasound images effectively contributed to the preoperative diagnosis of true and pseudo gallbladder polyps and may be valuable in their clinical management.

Study Processes and Applications of Ultrasomics in Precision Medicine

Ultrasomics is the science of transforming digitally encrypted medical ultrasound images that hold information related to tumor pathophysiology into mineable high-dimensional data. Ultrasomics data have the potential to uncover disease characteristics that are not found with the naked eye. The task of ultrasomics is to quantify the state of diseases using distinctive imaging algorithms and thereby provide valuable information for personalized medicine. Ultrasomics is a powerful tool in oncology but can also be applied to other medical problems for which a disease is imaged. To date there is no comprehensive review focusing on ultrasomics. Here, we describe how ultrasomics works and its capability in diagnosing disease in different organs, including breast, liver, and thyroid. Its pitfalls, challenges and opportunities are also discussed.

MR image-based radiomics to differentiate type Ι and type ΙΙ epithelial ovarian cancers

OBJECTIVES

Epithelial ovarian cancers (EOC) can be divided into type I and type II according to etiology and prognosis. Accurate subtype differentiation can substantially impact patient management. In this study, we aimed to construct an MR image-based radiomics model to differentiate between type I and type II EOC.

METHODS

In this multicenter retrospective study, a total of 294 EOC patients from January 2010 to February 2019 were enrolled. Quantitative MR imaging features were extracted from the following axial sequences: T2WI FS, DWI, ADC, and CE-T1WI. A combined model was constructed based on the combination of these four MR sequences. The diagnostic performance was evaluated by ROC-AUC. In addition, an occlusion test was carried out to identify the most critical region for EOC differentiation.

RESULTS

The combined radiomics model exhibited superior diagnostic capability over all four single-parametric radiomics models, both in internal and external validation cohorts (AUC of 0.806 and 0.847, respectively). The occlusion test revealed that the most critical region for differential diagnosis was the border zone between the solid and cystic components, or the less compact areas of solid component on direct visual inspection.

CONCLUSIONS

MR image-based radiomics modeling can differentiate between type I and type II EOC and identify the most critical region for differential diagnosis.

KEY POINTS

• Combined radiomics models exhibited superior diagnostic capability over all four single-parametric radiomics models, both in internal and external validation cohorts (AUC of 0.834 and 0.847, respectively). • The occlusion test revealed that the most crucial region for differentiating type Ι and type ΙΙ EOC was the border zone between the solid and cystic components, or the less compact areas of solid component on direct visual inspection on T2WI FS. • The light-combined model (constructed by T2WI FS, DWI, and ADC sequences) can be used for patients who are not suitable for contrast agent use.

Dual-mode ultrasound radiomics and intrinsic imaging phenotypes for diagnosis of lymph node lesions

Background

The ultrasonic diagnosis of lymph node lesions is usually based on a small number of subjective visual features from a single ultrasonic modality, which limits diagnostic accuracy. Therefore, our study aimed to propose a computerized method for using dual-mode ultrasound radiomics and the intrinsic imaging phenotypes for accurately differentiating benign, lymphomatous, and metastatic lymph nodes.

Methods

A total of 543 lymph nodes from 538 patients were examined with both B-mode ultrasonography and elastography. The data set was randomly divided into a training set of 407 nodes and a validation set of 136 nodes. First, we extracted 430 radiomic features from dual-mode images. Then, we combined the least absolute shrinkage and selection operator with the analysis of variance to select several typical features. We retrieved the intrinsic imaging phenotypes by using a hierarchical clustering of all radiomics features, and we integrated the phenotypes with the selected features for the classification of benign, lymphomatous, and metastatic nodes.

Results

The areas under the receiver operating characteristic curves (AUCs) on the validation set were 0.960 for benign vs. lymphomatous, 0.716 for benign vs. metastatic, 0.933 for lymphomatous vs. metastatic, and 0.856 for benign vs. malignant.

Conclusions

The radiomics features and intrinsic imaging phenotypes derived from the dual-mode ultrasound can capture the distinctions between benign, lymphomatous, and metastatic nodes and are valuable in node differentiation.

Advances in diagnostic application of ultrasomics in liver lesions

With the progress of medical technology in recent years, radiomics has been rapidly developed and widely used. Ultrasomics, as a branch of radiomics, is gradually applied to liver cancer, breast cancer, and other fields, and some research results have been acknowledged by clinicians. In the study of liver lesions, ultrasound is a vital diagnostic imaging method, but it also has limitations. For example, its performance is inferior to computed tomography or magnetic resonance imaging with regard to the diagnostic specificity for benignity and malignancy. The introduction and progress of ultrasomics provide new methods and ideas that could improve the ability to identify benignity or malignancy of liver lesions, tumor stage, and prognosis of the disease.

Radiomics in breast cancer classification and prediction

Breast Cancer (BC) is the common form of cancer in women. Its diagnosis and screening are usually performed through different imaging modalities such as mammography, magnetic resonance imaging and ultrasound. However, mammography and ultrasound-imaging techniques have limited sensitivity and specificity both in identifying lesions and in differentiating malign from benign lesions, especially in presence of dense breast parenchyma. Due to the higher resolution of magnetic resonance images, MRI represents the method with the higher specificity and sensitivity among all the available tools, in both lesions' identification and diagnosis. However, especially for diagnosis, even MRI has limitations that are only partially solved if combined with mammography. Unfortunately, due to the limits of all these imaging tools, in order to have a certain diagnosis, patients often receive painful and costly bioptics procedures. In this context, several computational approaches have been developed to increase sensitivity, while maintaining the same specificity, in BC diagnosis and screening. Amongst these, radiomics has been increasingly gaining ground in oncology to improve cancer diagnosis, prognosis and treatment. Radiomics derives multiple quantitative features from single or multiple medical imaging modalities, highlighting image traits which are not visible to the naked eye and hence significantly augmenting the discriminatory and predictive potential of medical imaging. This review article aims to summarize the state of the art in radiomics-based BC research. The dominating evidence extracted from the literature points towards a high potential of radiomics in disentangling malignant from benign breast lesions, classifying BC types and grades and also in predicting treatment response and recurrence risk. In the era of personalized medicine, radiomics has the potential to improve diagnosis, prognosis, prediction, monitoring, image-based intervention, and assessment of therapeutic response in BC.

A Radiomics Approach for the Classification of Fibroepithelial Lesions on Breast Ultrasonography

A radiomics-based classifier to distinguish phyllodes tumor and fibroadenoma on gray-scale breast ultrasonography was developed and validated. A total of 93 radiomics features were extracted from representative transverse plane ultrasound images of 182 fibroepithelial lesions initially diagnosed by core needle biopsy. High-throughput radiomics features were selected using the intra-class correlation coefficient between two radiologist readers and the Least Absolute Shrinkage and Selection Operator regression through 10-fold cross-validation. When applied to the validation set, the radiomics classifier for the differentiation of phyllodes tumors and benign/fibroadenomas achieved an area under the receiver operating characteristic curve of 0.765 (95% confidence interval [CI]: 0.597-0.888) with an accuracy of 0.703 (sensitivity: 0.857; specificity: 0.5). Our radiomics signature-based classifier may help predict phyllodes tumors among fibroepithelial lesions on breast ultrasonography.

Multimodal feature learning and fusion on B-mode ultrasonography and sonoelastography using point-wise gated deep networks for prostate cancer diagnosis

B-mode ultrasonography and sonoelastography are used in the clinical diagnosis of prostate cancer (PCa). A combination of the two ultrasound (US) modalities using computer aid may be helpful for improving the diagnostic performance. A technique for computer-aided diagnosis (CAD) of PCa is presented based on multimodal US. Firstly, quantitative features are extracted from both B-mode US images and sonoelastograms, including intensity statistics, regional percentile features, gray-level co-occurrence matrix (GLCM) texture features and binary texture features. Secondly, a deep network named PGBM-RBM2 is proposed to learn and fuse multimodal features, which is composed of the point-wise gated Boltzmann machine (PGBM) and two layers of the restricted Boltzmann machines (RBMs). Finally, the support vector machine (SVM) is used for prostatic disease classification. Experimental evaluation was conducted on 313 multimodal US images of the prostate from 103 patients with prostatic diseases (47 malignant and 56 benign). Under five-fold cross-validation, the classification sensitivity, specificity, accuracy, Youden's index and area under the receiver operating characteristic (ROC) curve with the PGBM-RBM2 were 87.0%, 88.8%, 87.9%, 75.8% and 0.851, respectively. The results demonstrate that multimodal feature learning and fusion using the PGBM-RBM2 can assist in the diagnosis of PCa. This deep network is expected to be useful in the clinical diagnosis of PCa.

Differential Diagnosis of Benign and Malignant Thyroid Nodules Using Deep Learning Radiomics of Thyroid Ultrasound Images

PURPOSE

We aimed to propose a highly automatic and objective model named deep learning Radiomics of thyroid (DLRT) for the differential diagnosis of benign and malignant thyroid nodules from ultrasound (US) images.

METHODS

We retrospectively enrolled and finally include US images and fine-needle aspiration biopsies from 1734 patients with 1750 thyroid nodules. A basic convolutional neural network (CNN) model, a transfer learning (TL) model, and a newly designed model named deep learning Radiomics of thyroid (DLRT) were used for the investigation. Their diagnostic accuracy was further compared with human observers (one senior and one junior US radiologist). Moreover, the robustness of DLRT over different US instruments was also validated. Analysis of receiver operating characteristic (ROC) curves were performed to calculate optimal area under it (AUC) for benign and malignant nodules. One observer helped to delineate the nodules.

RESULTS

AUCs of DLRT were 0.96 (95% confidence interval [CI]: 0.94-0.98), 0.95 (95% confidence interval [CI]: 0.93-0.97) and 0.97 (95% confidence interval [CI]: 0.95-0.99) in the training, internal and external validation cohort, respectively, which were significantly better than other deep learning models (P < 0.01) and human observers (P < 0.001). No significant difference was found when applying DLRT on thyroid US images acquired from different US instruments.

CONCLUSIONS

DLRT shows the best overall performance comparing with other deep learning models and human observers. It holds great promise for improving the differential diagnosis of benign and malignant thyroid nodules.

Online Transfer Learning for Differential Diagnosis of Benign and Malignant Thyroid Nodules With Ultrasound Images

OBJECTIVE

We aimed to propose a highly automatic and objective model named online transfer learning (OTL) for the differential diagnosis of benign and malignant thyroid nodules from ultrasound (US) images.

METHODS

The OTL mothed combined the strategy of transfer learning and online learning. Two datasets (1750 thyroid nodules with 1078 benign and 672 malignant nodules, and 3852 thyroid nodules with 3213 benign and 639 malignant nodules) were collected to develop the model. The diagnostic accuracy was also compared with VGG-16 based transfer learning model and different input images based model. Analysis of receiver operating characteristic (ROC) curves were performed to calculate optimal area under it (AUC) for benign and malignant nodules.

RESULTS

AUC, sensitivity and specificity of OTL were 0.98 (95% confidence interval [CI]: 0.97-0.99), 98.7% (95% confidence interval [CI]: 97.8%-99.6%) and 98.8% (95% confidence interval [CI]: 97.9%-99.7%) in the final online learning step, which was significantly better than other deep learning models (P < 0.01).

CONCLUSION

OTL model shows the best overall performance comparing with other deep learning models. The model holds a good potential for improving the overall diagnostic efficacy in thyroid nodule US examinations.

SIGNIFICANCE

The proposed OTL model could be seamlessly integrated into the conventional work-flow of thyroid nodule US examinations.

Radiomics With Attribute Bagging for Breast Tumor Classification Using Multimodal Ultrasound Images

OBJECTIVES

We aimed to develop radiomics with attribute bagging, which leverages multimodal ultrasound (US) images to improve the classification accuracy of breast tumors.

METHODS

A retrospective study was conducted. B-mode US, shear wave elastographic, and contrast-enhanced US images of 178 patients with 181 tumors (67 malignant and 114 benign) were included. Radiomics with attribute bagging consisted of extraction of 1226 radiomic features and analysis of them with attribute bagging. Histologic examination results acted as the reference standard. Radiomics with several feature selection algorithms were used for comparison. Cross-validation and a holdout test were performed to evaluate their performances.

RESULTS

The accuracy, sensitivity, specificity, and area under the receiver operating characteristic curve of radiomics with attribute bagging with the multimodal US images were 84.12%, 92.86%, 78.80%, and 0.919, respectively, exceeding all the comparison methods.

CONCLUSIONS

Radiomics with attribute bagging combined with multimodal US images has the potential to be used for accurate diagnosis of breast tumors in the clinic.

Ultrasound-Based Diagnosis of Breast Tumor with Parameter Transfer Multilayer Kernel Extreme Learning Machine

B-mode ultrasound (BUS) imaging is widely used for diagnosis of breast tumors. In recent years, another ultrasound modality, namely ultrasound elastography (UE), has also been successfully applied in clinical practice. Although the combination of bimodal ultrasound imaging can improve the diagnosis accuracy for breast tumors, the single-modal ultrasound-based diagnosis is more popular in clinical practice, especially in the rural areas. Since transfer learning (TL) can transfer knowledge from a source domain to a target domain to help train more effective model, we propose to develop a single-modal ultrasound-based computer-aided diagnosis (CAD) with the transferred information from an additional modality in this work. We propose a projective model (PM) based multilayer kernel extreme learning machine (ML-KELM-PM) algorithm, which performs the parameter transfer approach in classifier to perform TL in CAD. The experimental results on a bimodal ultrasound image dataset show that the proposed ML-KELM-PM algorithm outperforms all the compared algorithms for the single-modal ultrasound-based CAD for breast tumors.

Quantitative Multiparametric Breast Ultrasound: Application of Contrast-Enhanced Ultrasound and Elastography Leads to an Improved Differentiation of Benign and Malignant Lesions

OBJECTIVES

The aim of this study was to evaluate breast multiparametric ultrasound (mpUS) and its potential to reduce unnecessary breast biopsies with 1, 2, or 3 additional quantitative parameters (Doppler, elastography, and contrast-enhanced ultrasound [CEUS]) to B-mode and investigate possible variations with different reader experience.

MATERIALS AND METHODS

This prospective study included 124 women (age range, 18-82 years; mean, 52 years), each with 1 new breast lesion, scheduled for ultrasound-guided biopsy between October 2015 and September 2016. Each lesion was examined with B-mode, elastography (Virtual Touch IQ [VTIQ]), Doppler, and CEUS, and different quantitative parameters were recorded for each modality. Four readers (2 experienced breast radiologists and 2 in-training) independently evaluated B-mode images of each lesion and assigned a BI-RADS (Breast Imaging Reporting and Data System) score. Using the area under the receiver operating characteristic curve (AUC), the most accurate quantitative parameter for each modality was chosen. These were then combined with the BI-RADS scores of all readers. Descriptive statistics and AUC were used to evaluate the diagnostic performance of mpUS.

RESULTS

Sixty-five lesions were malignant. MpUS with B-mode and 2 additional quantitative parameters (VTIQ and CEUS or Doppler) showed the highest diagnostic performance for all readers (averaged AUCs, 0.812-0.789 respectively vs 0.683 for B-mode, P = 0.0001). Both combinations significantly reduced the number of false-positive findings up to 46.9% (P < 0.0001).

CONCLUSIONS

Quantitative mpUS with 2 different triple assessment modalities (B-mode, VTIQ elastography, CEUS, or Doppler) shows the best diagnostic performance for breast cancer diagnosis and leads to a significant reduction of false-positive biopsy recommendations, for both experienced and inexperienced readers.

Preoperative Prediction of Microvascular Invasion of Hepatocellular Carcinoma: Radiomics Algorithm Based on Ultrasound Original Radio Frequency Signals

Background: To evaluate the accuracy of radiomics algorithm based on original radio frequency (ORF) signals for prospective prediction of microvascular invasion (MVI) in hepatocellular carcinoma (HCC) lesions.

Methods: In this prospective study, we enrolled 42 inpatients diagnosed with HCC from January 2018 to December 2018. All HCC lesions were proved by surgical resection and histopathology results, including 21 lesions with MVI. Ultrasound ORF data and grayscale ultrasound images of HCC lesions were collected before operation for further radiomics analysis. Three ultrasound feature maps were calculated using signal analysis and processing (SAP) technology in first feature extraction. The diagnostic accuracy of model based on ORF signals was compared with the model based on grayscale ultrasound images.

Results: A total of 1,050 radiomics features were extracted from ORF signals of each HCC lesion. The performance of MVI prediction model based on ORF was better than those based on grayscale ultrasound images. The best area under curve, accuracy, sensitivity, and specificity of ultrasound radiomics in prediction of MVI were 95.01, 92.86, 85.71, and 100%, respectively.

Conclusions: Radiomics algorithm based on ultrasound ORF data combined with SAP technology can effectively predict MVI, which has potential clinical application value for non-invasively preoperative prediction of MVI in HCC patients.

Radiomics in stratification of pancreatic cystic lesions: Machine learning in action

Pancreatic cystic lesions (PCLs) are well-known precursors of pancreatic cancer. Their diagnosis can be challenging as their behavior varies from benign to malignant disease. Precise and timely management of malignant pancreatic cysts might prevent transformation to pancreatic cancer. However, the current consensus guidelines, which rely on standard imaging features to predict cyst malignancy potential, are conflicting and unclear. This has led to an increased interest in radiomics, a high-throughput extraction of comprehensible data from standard of care images. Radiomics can be used as a diagnostic and prognostic tool in personalized medicine. It utilizes quantitative image analysis to extract features in conjunction with machine learning and artificial intelligence (AI) methods like support vector machines, random forest, and convolutional neural network for feature selection and classification. Selected features can then serve as imaging biomarkers to predict high-risk PCLs. Radiomics studies conducted heretofore on PCLs have shown promising results. This cost-effective approach would help us to differentiate benign PCLs from malignant ones and potentially guide clinical decision-making leading to better utilization of healthcare resources. In this review, we discuss the process of radiomics, its myriad applications such as diagnosis, prognosis, and prediction of therapy response. We also discuss the outcomes of studies involving radiomic analysis of PCLs and pancreatic cancer, and challenges associated with this novel field along with possible solutions. Although these studies highlight the potential benefit of radiomics in the prevention and optimal treatment of pancreatic cancer, further studies are warranted before incorporating radiomics into the clinical decision support system.

Multiple Kernel Learning Based Classification of Parkinson's Disease With Multi-Modal Transcranial Sonography

Parkinson's Disease (PD) is the most common motor neurodegenerative disease in elderly population. Transcranial sonography (TCS) has become a popular imaging tool for diagnosis of PD in clinical practice. Moreover, several pioneering work have developed the computer-aided diagnosis (CAD) for PD with the transcranial B-mode sonography (TBS). It is worth noting that TCS not only has the TBS modality, but also can image the blood flow of major cerebral arteries, which is named transcranial Doppler sonography (TDS). TDS also has been applied to evaluate PD patients with orthostatic hypotension. However, the TDS-based CAD for PD has not been investigated. Since TBS and TDS provide the complementary structural and functional information about brain, it is feasible to develop a multi-modal TCS-based CAD for PD by combining both TBS and TDS. Therefore, in this work, we propose a multiple kernel learning (MKL) based CAD for PD with multi-modal TCS imaging. Particularly, the statistical and texture features are extracted from the midbrain region from TBS images, and the features about blood flow are calculated from the spectrum curves in TDS. The multi-modal features are then fed to a MKL classifier for classification of PD. The experimental results show that the multi-modal TCS-based method outperforms both the single-modal TBS- and TDS-based algorithm, which suggests the feasibility and effectiveness of combining TBS and TDS for diagnosis of PD.