Introduction to Radiomics

Radiomics Study for Discriminating Second Primary Lung Cancers From Pulmonary Metastases in Pulmonary Solid Lesions

BACKGROUND

The objective of this study was to assess the value of quantitative radiomics features in discriminating second primary lung cancers (SPLCs) from pulmonary metastases (PMs).

METHODS

This retrospective study enrolled 252 malignant pulmonary nodules with histopathologically confirmed SPLCs or PMs and randomly assigned them to a training or validation cohort. Clinical data were collected from the electronic medical records system. The imaging and radiomics features of each nodule were extracted from CT images.

RESULTS

A rad-score was generated from the training cohort using the least absolute shrinkage and selection operator regression. A clinical and radiographic model was constructed using the clinical and imaging features selected by univariate and multivariate regression. A nomogram composed of clinical-radiographic factors and a rad-score were developed to validate the discriminative ability. The rad-scores differed significantly between the SPLC and PM groups. Sixteen radiomics features and four clinical-radiographic features were selected to build the final model to differentiate between SPLCs and PMs. The comprehensive clinical radiographic-radiomics model demonstrated good discriminative capacity with an area under the curve of the receiver operating characteristic curve of 0.9421 and 0.9041 in the respective training and validation cohorts. The decision curve analysis demonstrated that the comprehensive model showed a higher clinical value than the model without the rad-score.

CONCLUSION

The proposed model based on clinical data, imaging features, and radiomics features could accurately discriminate SPLCs from PMs. The model thus has the potential to support clinicians in improving decision-making in a noninvasive manner.

Preoperative prediction of residual back pain after vertebral augmentation for osteoporotic vertebral compression fractures: Initial application of a radiomics score based nomogram

BACKGROUND

Most patients with osteoporotic vertebral compression fracture (OVCF) obtain pain relief after vertebral augmentation, but some will experience residual back pain (RBP) after surgery. Although several risk factors of RBP have been reported, it is still difficult to estimate the risk of RBP preoperatively. Radiomics is helpful for disease diagnosis and outcome prediction by establishing complementary relationships between human-recognizable and computer-extracted features. However, musculoskeletal radiomics investigations are less frequently reported.

OBJECTIVE

This study aims to establish a radiomics score (rad-score) based nomogram for the preoperative prediction of RBP in OVCF patients.

METHODS

The training cohort of 731 OVCF patients was used for nomogram development, and the validation cohort was utilized for performance test. RBP was determined as the score of visual analogue scale ≥ 4 at both 3 and 30 days following surgery. After normalization, the RBP-related radiomics features were selected to create rad-scores. These rad-scores, along with the RBP predictors initially identified by univariate analyses, were included in the multivariate analysis to establish a nomogram for the assessment of the RBP risk in OVCF patients preoperatively.

RESULTS

A total of 81 patients (11.2%) developed RBP postoperatively. We finally selected 8 radiomics features from 1316 features extracted from each segmented image to determine the rad-score. Multivariate analysis revealed that the rad-score plus bone mineral density, intravertebral cleft, and thoracolumbar fascia injury were independent factors of RBP. Our nomograms based on these factors demonstrated good discrimination, calibration, and clinical utility in both training and validation cohorts. Furthermore, it achieved better performance than the rad-score itself, as well as the nomogram only incorporating regular features.

CONCLUSION

We developed and validated a nomogram incorporating the rad-score and regular features for preoperative prediction of the RBP risk in OVCF patients, which contributed to improved surgical outcomes and patient satisfaction.

Radiomics in prostate cancer: an up-to-date review

Prostate cancer (PCa) is the most common worldwide diagnosed malignancy in male population. The diagnosis, the identification of aggressive disease, and the post-treatment follow-up needs a more comprehensive and holistic approach. Radiomics is the extraction and interpretation of images phenotypes in a quantitative manner. Radiomics may give an advantage through advancements in imaging modalities and through the potential power of artificial intelligence techniques by translating those features into clinical outcome prediction. This article gives an overview on the current evidence of methodology and reviews the available literature on radiomics in PCa patients, highlighting its potential for personalized treatment and future applications.

The value of radiomics to predict abnormal bone mass in type 2 diabetes mellitus patients based on CT imaging for paravertebral muscles

OBJECTIVE

To investigate the value of CT imaging features of paravertebral muscles in predicting abnormal bone mass in patients with type 2 diabetes mellitus.

METHODS

The clinical and QCT data of 149 patients with type 2 diabetes mellitus were collected retrospectively. Patients were randomly divided into the training group (n = 90) and the validation group (n = 49). The radiologic model and Nomogram model were established by multivariate Logistic regression analysis. Predictive performance was evaluated using receiver operating characteristic (ROC) curves.

RESULTS

A total of 829 features were extracted from CT images of paravertebral muscles, and 12 optimal predictive features were obtained by the mRMR and Lasso feature selection methods. The radiomics model can better predict bone abnormality in type 2 diabetes mellitus, and the (Area Under Curve) AUC values of the training group and the validation group were 0.94(95% CI, 0.90-0.99) and 0.90(95% CI, 0.82-0.98). The combined Nomogram model, based on radiomics and clinical characteristics (vertebral CT values), showed better predictive efficacy with an AUC values of 0.97(95% CI, 0.94-1.00) in the training group and 0.95(95% CI, 0.90-1.00) in the validation group, compared with the clinical model.

CONCLUSION

The combination of Nomogram model and radiomics-clinical features of paravertebral muscles has a good predictive value for abnormal bone mass in patients with type 2 diabetes mellitus.

Machine Learning-Based CT Radiomics Method for Identifying the Stage of Wilms Tumor in Children

PURPOSE

To develop and validate a machine learning-based CT radiomics method for preoperatively predicting the stages (stage I and non-stage I) of Wilms tumor (WT) in pediatric patients.

METHODS

A total of 118 patients with WT, who underwent contrast-enhanced computed tomography (CT) scans in our center between 2014 and 2021, were studied retrospectively and divided into two groups: stage I and non-stage I disease. Patients were randomly divided into training cohorts (n = 94) and test cohorts (n = 24). A total of 1,781 radiomic features from seven feature classes were extracted from preoperative portal venous-phase images of abdominal CT. Synthetic Minority Over-Sampling Technique (SMOTE) was used to handle imbalanced datasets, followed by a t-test and Least Absolute Shrinkage and Selection Operator (LASSO) regularization for feature selection. Support Vector Machine (SVM) was deployed using the selected informative features to develop the predicting model. The performance of the model was evaluated according to its accuracy, sensitivity, and specificity. The receiver operating characteristic curve (ROC) and the area under the ROC curve (AUC) was also arranged to assess the model performance.

RESULTS

The SVM model was fitted with 15 radiomic features obtained by t-test and LASSO concerning WT staging in the training dataset and demonstrated favorable performance in the testing dataset. Cross-validated AUC on the training dataset was 0.79 with a 95 percent confidence interval (CI) of 0.773-0.815 and a coefficient of variation of 3.76%, while AUC on the test dataset was 0.81, and accuracy, sensitivity, and specificity were 0.79, 0.87, and 0.69, respectively.

CONCLUSIONS

The machine learning model of SVM based on radiomic features extracted from CT images accurately predicted WT stage I and non-stage I disease in pediatric patients preoperatively, which provided a rapid and non-invasive way for investigation of WT stages.

Radiomics analysis based on lumbar spine CT to detect osteoporosis

OBJECTIVES

Undiagnosed osteoporosis may lead to severe complications after spinal surgery. This study aimed to construct and validate a radiomic signature based on CT scans to screen for lumbar spine osteoporosis.

METHODS

Using a stratified random sample method, 386 vertebral bodies were randomly divided into a training set (n = 270) and a test set (n = 116). A total of 1040 radiomics features were automatically retracted from lumbar spine CT scans using the 3D slicer pyradiomics module, and a radiomic signature was created. The sensitivity, specificity, accuracy, and area under the receiver operating characteristic curve (AUC) of the Hounsfield and radiomics signature models were calculated. The AUCs of the two models were compared using the DeLong test. Their clinical usefulness was assessed using a decision curve analysis.

RESULTS

Twelve features were chosen to establish the radiomic signature. The AUCs of the radiomics signature and Hounsfield models were 0.96 and 0.88 in the training set and 0.92 and 0.84 in the test set, respectively. According to the DeLong test, the AUCs of the two models were significantly different (p < 0.05). The radiomics signature model indicated a higher overall net benefit than the Hounsfield model, as determined by decision curve analysis.

CONCLUSIONS

The CT-based radiomic signature can differentiate patients with/without osteoporosis prior to lumbar spinal surgery. Without additional medical cost and radiation exposure, the radiomics method may provide valuable information facilitating surgical decision-making.

KEY POINTS

• The goal of the study was to evaluate the efficacy of a radiomics signature model based on routine preoperative lumbar spine CT scans in screening osteoporosis. • The radiomics signature model demonstrated excellent prediction performance in both the training and test sets. • This radiomics method may provide valuable information and facilitate surgical decision-making without additional medical costs and radiation exposure.

Combined model of radiomics and clinical features for differentiating pneumonic-type mucinous adenocarcinoma from lobar pneumonia: An exploratory study

PURPOSE

The purpose of this study was to distinguish pneumonic-type mucinous adenocarcinoma (PTMA) from lobar pneumonia (LP) by pre-treatment CT radiological and clinical or radiological parameters.

METHODS

A total of 199 patients (patients diagnosed with LP = 138, patients diagnosed with PTMA = 61) were retrospectively evaluated and assigned to either the training cohort (n = 140) or the validation cohort (n = 59). Radiomics features were extracted from chest CT plain images. Multivariate logistic regression analysis was conducted to develop a radiomics model and a nomogram model, and their clinical utility was assessed. The performance of the constructed models was assessed with the receiver operating characteristic (ROC) curve and the area under the curve (AUC). The clinical application value of the models was comprehensively evaluated using decision curve analysis (DCA).

RESULTS

The radiomics signature, consisting of 14 selected radiomics features, showed excellent performance in distinguishing between PTMA and LP, with an AUC of 0.90 (95% CI, 0.83-0.96) in the training cohort and 0.88 (95% CI, 0.79-0.97) in the validation cohort. A nomogram model was developed based on the radiomics signature and clinical features. It had a powerful discriminative ability, with the highest AUC values of 0.94 (95% CI, 0.90-0.98) and 0.91 (95% CI, 0.84-0.99) in the training cohort and validation cohort, respectively, which were significantly superior to the clinical model alone. There were no significant differences in calibration curves from Hosmer-Lemeshow tests between training and validation cohorts (p = 0.183 and p = 0.218), which indicated the good performance of the nomogram model. DCA indicated that the nomogram model exhibited better performance than the clinical model.

CONCLUSIONS

The nomogram model based on radiomics signatures of CT images and clinical risk factors could help to differentiate PTMA from LP, which can provide appropriate therapy decision support for clinicians, especially in situations where differential diagnosis is difficult.

Artificial Intelligence in Lymphoma PET Imaging:: A Scoping Review (Current Trends and Future Directions)

Malignant lymphomas are a family of heterogenous disorders caused by clonal proliferation of lymphocytes. 18F-FDG-PET has proven to provide essential information for accurate quantification of disease burden, treatment response evaluation, and prognostication. However, manual delineation of hypermetabolic lesions is often a time-consuming and impractical task. Applications of artificial intelligence (AI) may provide solutions to overcome this challenge. Beyond segmentation and detection of lesions, AI could enhance tumor characterization and heterogeneity quantification, as well as treatment response prediction and recurrence risk stratification. In this scoping review, we have systematically mapped and discussed the current applications of AI (such as detection, classification, segmentation as well as the prediction and prognostication) in lymphoma PET.

Ultrasound images-based deep learning radiomics nomogram for preoperative prediction of RET rearrangement in papillary thyroid carcinoma

PURPOSE

To create an ultrasound -based deep learning radiomics nomogram (DLRN) for preoperatively predicting the presence of RET rearrangement among patients with papillary thyroid carcinoma (PTC).

METHODS

We retrospectively enrolled 650 patients with PTC. Patients were divided into the RET/PTC rearrangement group (n = 103) and the non-RET/PTC rearrangement group (n = 547). Radiomics features were extracted based on hand-crafted features from the ultrasound images, and deep learning networks were used to extract deep transfer learning features. The least absolute shrinkage and selection operator regression was applied to select the features of nonzero coefficients from radiomics and deep transfer learning features; then, we established the deep learning radiomics signature. DLRN was constructed using a logistic regression algorithm by combining clinical and deep learning radiomics signatures. The prediction performance was evaluated using the receiver operating characteristic curve, calibration curve, and decision curve analysis.

RESULTS

Comparing the effectiveness of the models by linking the area under the receiver operating characteristic curve of each model, we found that the area under the curve of DLRN could reach 0.9545 (95% confidence interval: 0.9133-0.9558) in the test cohort and 0.9396 (95% confidence interval: 0.9185-0.9607) in the training cohort, indicating that the model has an excellent performance in predicting RET rearrangement in PTC. The decision curve analysis demonstrated that the combined model was clinically useful.

CONCLUSION

The novel ultrasonic-based DLRN has an important clinical value for predicting RET rearrangement in PTC. It can provide physicians with a preoperative non-invasive primary screening method for RET rearrangement diagnosis, thus facilitating targeted patients with purposeful molecular sequencing to avoid unnecessary medical investment and improve treatment outcomes.

Predictive Value of a Combined Model Based on Pre-Treatment and Mid-Treatment MRI-Radiomics for Disease Progression or Death in Locally Advanced Nasopharyngeal Carcinoma

Purpose

A combined model was established based on the MRI-radiomics of pre- and mid-treatment to assess the risk of disease progression or death in locally advanced nasopharyngeal carcinoma.

Materials and Methods

A total of 243 patients were analyzed. We extracted 10,400 radiomics features from the primary nasopharyngeal tumors and largest metastatic lymph nodes on the axial contrast-enhanced T1 weighted and T2 weighted in pre- and mid-treatment MRI, respectively. We used the SMOTE algorithm, center and scale and box-cox, Pearson correlation coefficient, and LASSO regression to construct the pre- and mid-treatment MRI-radiomics prediction model, respectively, and the risk scores named P score and M score were calculated. Finally, univariate and multivariate analyses were used for P score, M score, and clinical data to build the combined model and grouped the patients into two risk levels, namely, high and low.

Result

A combined model of pre- and mid-treatment MRI-radiomics successfully categorized patients into high- and low-risk groups. The log-rank test showed that the high- and low-risk groups had good prognostic performance in PFS (P<0.0001, HR: 19.71, 95% CI: 12.77–30.41), which was better than TNM stage (P=0.004, HR:1.913, 95% CI:1.250–2.926), and also had an excellent predictive effect in LRFS, DMFS, and OS.

Conclusion

Risk grouping of LA-NPC using a combined model of pre- and mid-treatment MRI-radiomics can better predict disease progression or death.

Methodological quality of machine learning-based quantitative imaging analysis studies in esophageal cancer: a systematic review of clinical outcome prediction after concurrent chemoradiotherapy

PURPOSE

Studies based on machine learning-based quantitative imaging techniques have gained much interest in cancer research. The aim of this review is to critically appraise the existing machine learning-based quantitative imaging analysis studies predicting outcomes of esophageal cancer after concurrent chemoradiotherapy in accordance with PRISMA guidelines.

METHODS

A systematic review was conducted in accordance with PRISMA guidelines. The citation search was performed via PubMed and Embase Ovid databases for literature published before April 2021. From each full-text article, study characteristics and model information were summarized. We proposed an appraisal matrix with 13 items to assess the methodological quality of each study based on recommended best-practices pertaining to quality.

RESULTS

Out of 244 identified records, 37 studies met the inclusion criteria. Study endpoints included prognosis, treatment response, and toxicity after concurrent chemoradiotherapy with reported discrimination metrics in validation datasets between 0.6 and 0.9, with wide variation in quality. A total of 30 studies published within the last 5 years were evaluated for methodological quality and we found 11 studies with at least 6 "good" item ratings.

CONCLUSION

A substantial number of studies lacked prospective registration, external validation, model calibration, and support for use in clinic. To further improve the predictive power of machine learning-based models and translate into real clinical applications in cancer research, appropriate methodologies, prospective registration, and multi-institution validation are recommended.

Development and Validation of a Nomogram Based on 18F-FDG PET/CT Radiomics to Predict the Overall Survival in Adult Hemophagocytic Lymphohistiocytosis

Purpose: Hemophagocytic lymphohistiocytosis (HLH) is a rare and severe disease with a poor prognosis. We aimed to determine if ¹⁸F-fluorodeoxyglucose (¹⁸F-FDG) PET/CT-derived radiomic features alone or combination with clinical parameters could predict survival in adult HLH.

Methods: This study included 70 adults with HLH (training cohort, n = 50; validation cohort, n = 20) who underwent pretherapeutic ¹⁸F-FDG PET/CT scans between August 2016 and June 2020. Radiomic features were extracted from the liver and spleen on CT and PET images. For evaluation of 6-month survival, the features exhibiting p < 0.1 in the univariate analysis between non-survivors and survivors were selected. The least absolute shrinkage and selection operator (LASSO) regression analysis was used to develop a radiomics score (Rad-score). A nomogram was built by the multivariate regression analysis to visualize the predictive model for 3-month, 6-month, and 1-year survival, while the performance and usefulness of the model were evaluated by calibration curves, the receiver operating characteristic (ROC) curves, and decision curves.

Results: The Rad-score was able to predict 6-month survival in adult HLH, with area under the ROC curves (AUCs) of 0.927 (95% CI: 0.878–0.974) and 0.869 (95% CI: 0.697–1.000) in the training and validation cohorts, respectively. The radiomics nomogram combining the Rad-score with the clinical parameters resulted in better performance for predicting 6-month survival than the clinical model or the Rad-score alone. Moreover, the nomogram displayed superior discrimination, calibration, and clinical usefulness in both the cohorts.

Conclusion: The newly developed Rad-score is a powerful predictor for overall survival (OS) in adults with HLH. The nomogram has great potential for predicting 3-month, 6-month, and 1-year survival, which may timely guide personalized treatments for adult HLH.

MRI Radiomics in Prostate Cancer: A Reliability Study

BACKGROUND

Radiomics can provide quantitative features from medical imaging that can be correlated to clinical endpoints. The challenges relevant to robustness of radiomics features have been analyzed by many researchers, as it seems to be influenced by acquisition and reconstruction protocols, as well as by the segmentation of the region of interest (ROI). Prostate cancer (PCa) represents a difficult playground for this technique, due to discrepancies in the identification of the cancer lesion and the heterogeneity of the acquisition protocols. The aim of this study was to investigate the reliability of radiomics in PCa magnetic resonance imaging (MRI).

METHODS

A homogeneous cohort of patients with a PSA rise that underwent multiparametric MRI imaging of the prostate before biopsy was tested in this study. All the patients were acquired with the same MRI scanner, with a standardized protocol. The identification and the contouring of the region of interest (ROI) of an MRI suspicious cancer lesion were done by two radiologists with great experience in prostate cancer (>10 years). After the segmentation, the texture features were extracted with LIFEx. Texture features were then tested with intraclass coefficient correlation (ICC) analysis to analyze the reliability of the segmentation.

RESULTS

Forty-four consecutive patients were included in the present analysis. In 26 patients (59.1%), the prostate biopsy confirmed the presence of prostate cancer, which was scored as Gleason 6 in 6 patients (13.6%), Gleason 3 + 4 in 8 patients (18.2%), and Gleason 4 + 3 in 12 patients (27.3%). The reliability analysis conversely showed poor reliability in the majority of the MRI acquisition (61% in T2, 89% in DWI50, 44% in DWI400, and 83% in DWI1,500), with ADC acquisition only showing better reliability (poor reliability in only 33% of the texture features).

CONCLUSIONS

The low ratio of reliability in a monoinstitutional homogeneous cohort represents a significant alarm bell for the application of MRI radiomics in the field of prostate cancer. More work is needed in a clinical setting to further study the potential of MRI radiomics in prostate cancer.

Prognostic and predictive value of radiomics features at MRI in nasopharyngeal carcinoma

PURPOSE

To explore the value of MRI-based radiomics features in predicting risk in disease progression for nasopharyngeal carcinoma (NPC).

METHODS

199 patients confirmed with NPC were retrospectively included and then divided into training and validation set using a hold-out validation (159: 40). Discriminative radiomic features were selected with a Wilcoxon signed-rank test from tumors and normal masticatory muscles of 37 NPC patients. LASSO Cox regression and Pearson correlation analysis were applied to further confirm the differential expression of the radiomic features in the training set. Using the multiple Cox regression model, we built a radiomic feature-based classifier, Rad-Score. The prognostic and predictive performance of Rad-Score was validated in the validation cohort and illustrated in all included 199 patients.

RESULTS

We identified 1832 differentially expressed radiomic features between tumors and normal tissue. Rad-Score was built based on one radiomic feature: CET1-w_wavelet.LLH_GLDM_Dependence-Entropy. Rad-Score showed a satisfactory performance to predict disease progression in NPC with an area under the curve (AUC) of 0.604, 0.732, 0.626 in the training, validation, and the combined cohort (all 199 patients included) respectively. Rad-Score improved risk stratification, and disease progression-free survival was significantly different between these groups in every cohort of patients (p = 0.044 or p < 0.01). Combining radiomics and clinical features, higher AUC was achieved of the prediction of 3-year disease progression-free survival (PFS) (AUC, 0.78) and 5-year disease PFS (AUC, 0.73), although there was no statistical difference.

CONCLUSION

The radiomics classifier, Rad-Score, was proven useful for pretreatment prognosis prediction and showed potential in risk stratification for NPC.

SUPPLEMENTARY INFORMATION

The online version contains supplementary material available at 10.1007/s12672-021-00460-3.

State of the Art and New Concepts in Giant Cell Tumor of Bone: Imaging Features and Tumor Characteristics

Giant cell tumor of bone (GCTB) is classified as an intermediate malignant tumor due to its locally aggressive behavior, burdened by high local recurrence rate. GCTB accounts for about 4-5% of all primary bone tumors and typically arises in the metaphysis and epiphyses of the long tubular bones. Mutation of gene H3F3A is at the basis of GCTB etiopathogenesis, and its immunohistochemical expression is a valuable method for practical diagnosis, even if new biomarkers have been identified for early diagnosis and for potential tumor recurrence prediction. In the era of computer-aided diagnosis, imaging plays a key role in the assessment of GCTB for surgical planning, patients' prognosis prediction and post treatment evaluation. Cystic changes, penetrating irregular margins and adjacent soft tissue invasion on preoperative Magnetic Resonance Imaging (MRI) have been associated with a higher rate of local recurrence. Distance from the tumor edge to the articular surface and thickness of unaffected cortical bone around the tumor should be evaluated on Computed Tomography (CT) as related to local recurrence. Main features associated with local recurrence after curettage are bone resorption around the graft or cement, soft tissue mass formation and expansile destruction of bone. A denosumab positive response is represented by a peripherical well-defined osteosclerosis around the lesion and intralesional ossification. Radiomics has proved to offer a valuable contribution in aiding GCTB pre-operative diagnosis through clinical-radiomics models based on CT scans and multiparametric MR imaging, possibly guiding the choice of a patient-tailored treatment. Moreover, radiomics models based on texture analysis demonstrated to be a promising alternative solution for the assessment of GCTB response to denosumab both on conventional radiography and CT since the quantitative variation of some radiomics features after therapy has been correlated with tumor response, suggesting they might facilitate disease monitoring during post-denosumab surveillance.

Feasibility on the Use of Radiomics Features of 11[C]-MET PET/CT in Central Nervous System Tumours: Preliminary Results on Potential Grading Discrimination Using a Machine Learning Model

Background/Aim: Nowadays, Machine Learning (ML) algorithms have demonstrated remarkable progress in image-recognition tasks and could be useful for the new concept of precision medicine in order to help physicians in the choice of therapeutic strategies for brain tumours. Previous data suggest that, in the central nervous system (CNS) tumours, amino acid PET may more accurately demarcate the active disease than paramagnetic enhanced MRI, which is currently the standard method of evaluation in brain tumours and helps in the assessment of disease grading, as a fundamental basis for proper clinical patient management. The aim of this study is to evaluate the feasibility of ML on 11[C]-MET PET/CT scan images and to propose a radiomics workflow using a machine-learning method to create a predictive model capable of discriminating between low-grade and high-grade CNS tumours. Materials and Methods: In this retrospective study, fifty-six patients affected by a primary brain tumour who underwent 11[C]-MET PET/CT were selected from January 2016 to December 2019. Pathological examination was available in all patients to confirm the diagnosis and grading of disease. PET/CT acquisition was performed after 10 min from the administration of 11C-Methionine (401–610 MBq) for a time acquisition of 15 min. 11[C]-MET PET/CT images were acquired using two scanners (24 patients on a Siemens scan and 32 patients on a GE scan). Then, LIFEx software was used to delineate brain tumours using two different semi-automatic and user-independent segmentation approaches and to extract 44 radiomics features for each segmentation. A novel mixed descriptive-inferential sequential approach was used to identify a subset of relevant features that correlate with the grading of disease confirmed by pathological examination and clinical outcome. Finally, a machine learning model based on discriminant analysis was used in the evaluation of grading prediction (low grade CNS vs. high-grade CNS) of 11[C]-MET PET/CT. Results: The proposed machine learning model based on (i) two semi-automatic and user-independent segmentation processes, (ii) an innovative feature selection and reduction process, and (iii) the discriminant analysis, showed good performance in the prediction of tumour grade when the volumetric segmentation was used for feature extraction. In this case, the proposed model obtained an accuracy of ~85% (AUC ~79%) in the subgroup of patients who underwent Siemens tomography scans, of 80.51% (AUC 65.73%) in patients who underwent GE tomography scans, and of 70.31% (AUC 64.13%) in the whole patients’ dataset (Siemens and GE scans). Conclusions: This preliminary study on the use of an ML model demonstrated to be feasible and able to select radiomics features of 11[C]-MET PET with potential value in prediction of grading of disease. Further studies are needed to improve radiomics algorithms to personalize predictive and prognostic models and potentially support the medical decision process.

Value of the Application of CE-MRI Radiomics and Machine Learning in Preoperative Prediction of Sentinel Lymph Node Metastasis in Breast Cancer

Purpose

To explore the value of machine learning model based on CE-MRI radiomic features in preoperative prediction of sentinel lymph node (SLN) metastasis of breast cancer.

Methods

The clinical, pathological and MRI data of 177 patients with pathologically confirmed breast cancer (81 with SLN positive and 96 with SLN negative) and underwent conventional DCE-MRI before surgery in the First Affiliated Hospital of Soochow University from January 2015 to May 2021 were analyzed retrospectively. The samples were randomly divided into the training set (n=123) and validation set (n= 54) according to the ratio of 7:3. The radiomic features were derived from DCE-MRI phase 2 images, and 1,316 original eigenvectors are normalized by maximum and minimum normalization. The optimal feature filter and selection operator (LASSO) algorithm were used to obtain the optimal features. Five machine learning models of Support Vector Machine, Random Forest, Logistic Regression, Gradient Boosting Decision Tree, and Decision Tree were constructed based on the selected features. Radiomics signature and independent risk factors were incorporated to build a combined model. The receiver operating characteristic curve and area under the curve were used to evaluate the performance of the above models, and the accuracy, sensitivity, and specificity were calculated.

Results

There is no significant difference between all clinical and histopathological variables in breast cancer patients with and without SLN metastasis (P >0.05), except tumor size and BI-RADS classification (P< 0.01). Thirteen features were obtained as optimal features for machine learning model construction. In the validation set, the AUC (0.86) of SVM was the highest among the five machine learning models. Meanwhile, the combined model showed better performance in sentinel lymph node metastasis (SLNM) prediction and achieved a higher AUC (0.88) in the validation set.

Conclusions

We revealed the clinical value of machine learning models established based on CE-MRI radiomic features, providing a highly accurate, non-invasive, and convenient method for preoperative prediction of SLNM in breast cancer patients.

Differentiation of Brain Abscess From Cystic Glioma Using Conventional MRI Based on Deep Transfer Learning Features and Hand-Crafted Radiomics Features

Objectives: To develop and validate the model for distinguishing brain abscess from cystic glioma by combining deep transfer learning (DTL) features and hand-crafted radiomics (HCR) features in conventional T1-weighted imaging (T1WI) and T2-weighted imaging (T2WI).

Methods: This single-center retrospective analysis involved 188 patients with pathologically proven brain abscess (102) or cystic glioma (86). One thousand DTL and 105 HCR features were extracted from the T1WI and T2WI of the patients. Three feature selection methods and four classifiers, such as k-nearest neighbors (KNN), random forest classifier (RFC), logistic regression (LR), and support vector machine (SVM), for distinguishing brain abscess from cystic glioma were compared. The best feature combination and classifier were chosen according to the quantitative metrics including area under the curve (AUC), Youden Index, and accuracy.

Results: In most cases, deep learning-based radiomics (DLR) features, i.e., DTL features combined with HCR features, contributed to a higher accuracy than HCR and DTL features alone for distinguishing brain abscesses from cystic gliomas. The AUC values of the model established, based on the DLR features in T2WI, were 0.86 (95% CI: 0.81, 0.91) in the training cohort and 0.85 (95% CI: 0.75, 0.95) in the test cohort, respectively.

Conclusions: The model established with the DLR features can distinguish brain abscess from cystic glioma efficiently, providing a useful, inexpensive, convenient, and non-invasive method for differential diagnosis. This is the first time that conventional MRI radiomics is applied to identify these diseases. Also, the combination of HCR and DTL features can lead to get impressive performance.

Machine Learning Algorithms in Neuroimaging: An Overview

Machine learning (ML) and artificial intelligence (AI) applications in the field of neuroimaging have been on the rise in recent years, and their clinical adoption is increasing worldwide. Deep learning (DL) is a field of ML that can be defined as a set of algorithms enabling a computer to be fed with raw data and progressively discover-through multiple layers of representation-more complex and abstract patterns in large data sets. The combination of ML and radiomics, namely the extraction of features from medical images, has proven valuable, too: Radiomic information can be used for enhanced image characterization and prognosis or outcome prediction. This chapter summarizes the basic concepts underlying ML application for neuroimaging and discusses technical aspects of the most promising algorithms, with a specific focus on Convolutional Neural Networks (CNNs) and Generative Adversarial Networks (GANs), in order to provide the readership with the fundamental theoretical tools to better understand ML in neuroimaging. Applications are highlighted from a practical standpoint in the last section of the chapter, including: image reconstruction and restoration, image synthesis and super-resolution, registration, segmentation, classification, and outcome prediction.

18F-FDG texture analysis predicts the pathological Fuhrman nuclear grade of clear cell renal cell carcinoma

PURPOSE

This article analyzes the image heterogeneity of clear cell renal cell carcinoma (ccRCC) based on positron emission tomography (PET) and positron emission tomography-computed tomography (PET/CT) texture parameters, and provides a new objective quantitative parameter for predicting pathological Fuhrman nuclear grading before surgery.

METHODS

A retrospective analysis was performed on preoperative PET/CT images of 49 patients whose surgical pathology was ccRCC, 27 of whom were low grade (Fuhrman I/II) and 22 of whom were high grade (Fuhrman III/IV). Radiological parameters and standard uptake value (SUV) indicators on PET and computed tomography (CT) images were extracted by using the LIFEx software package. The discriminative ability of each texture parameter was evaluated through receiver operating curve (ROC). Binary logistic regression analysis was used to screen the texture parameters with distinguishing and diagnostic capabilities and whose area under curve (AUC) > 0.5. DeLong's test was used to compare the AUCs of PET texture parameter model and PET/CT texture parameter model with traditional maximum standardized uptake value (SUVmax) model and the ratio of tumor SUVmax to liver SUVmean (SUL)model. In addition, the models with the larger AUCs among the SUV models and texture models were prospectively internally verified.

RESULTS

In the ROC curve analysis, the AUCs of SUVmax model, SUL model, PET texture parameter model, and PET/CT texture parameter model were 0.803, 0.819, 0.873, and 0.926, respectively. The prediction ability of PET texture parameter model or PET/CT texture parameter model was significantly better than SUVmax model (P = 0.017, P = 0.02), but it was not better than SUL model (P = 0.269, P = 0.053). In the prospective validation cohort, both the SUL model and the PET/CT texture parameter model had good predictive ability, and the AUCs of them were 0.727 and 0.792, respectively.

CONCLUSION

PET and PET/CT texture parameter models can improve the prediction ability of ccRCC Fuhrman nuclear grade; SUL model may be the more accurate and easiest way to predict ccRCC Fuhrman nuclear grade.

Emerging Applications of Radiomics in Neurological Disorders: A Review

Radiomics has achieved significant momentum in radiology research and can reveal image information invisible to radiologists' eyes. Radiomics first evolved for oncologic imaging. Oncologic applications (histopathology, tumor grading, gene mutation analysis, patient survival, and treatment response prediction) of radiomics are widespread. However, it is not limited to oncologic analysis, and any digital medical images can benefit from radiomics analysis. This article reviews the current literature on radiomics in non-oncologic, neurological disorders including ischemic strokes, hemorrhagic stroke, cerebral aneurysms, and demyelinating disorders.

Machine learning in the differentiation of follicular lymphoma from diffuse large B-cell lymphoma with radiomic [18F]FDG PET/CT features

BACKGROUND

One of the challenges in the management of patients with follicular lymphoma (FL) is the identification of individuals with histological transformation, most commonly into diffuse large B-cell lymphoma (DLBCL). [¹⁸F]FDG-PET/CT is used for staging of patients with lymphoma, but visual interpretation cannot reliably discern FL from DLBCL. This study evaluated whether radiomic features extracted from clinical baseline [¹⁸F]FDG PET/CT and analyzed by machine learning algorithms may help discriminate FL from DLBCL.

MATERIALS AND METHODS

Patients were selected based on confirmed histopathological diagnosis of primary FL (n=44) or DLBCL (n=76) and available [¹⁸F]FDG PET/CT with EARL reconstruction parameters within 6 months of diagnosis. Radiomic features were extracted from the volume of interest on co-registered [¹⁸F]FDG PET and CT images. Analysis of selected radiomic features was performed with machine learning classifiers based on logistic regression and tree-based ensemble classifiers (AdaBoosting, Gradient Boosting, and XG Boosting). The performance of radiomic features was compared with a SUV_max-based logistic regression model.

RESULTS

From the segmented lesions, 121 FL and 227 DLBCL lesions were included for radiomic feature extraction. In total, 79 radiomic features were extracted from the SUVmap, 51 from CT, and 6 shape features. Machine learning classifier Gradient Boosting achieved the best discrimination performance using 136 radiomic features (AUC of 0.86 and accuracy of 80%). SUV_max-based logistic regression model achieved an AUC of 0.79 and an accuracy of 70%. Gradient Boosting classifier had a significantly greater AUC and accuracy compared to the SUV_max-based logistic regression (p≤0.01).

CONCLUSION

Machine learning analysis of radiomic features may be of diagnostic value for discriminating FL from DLBCL tumor lesions, beyond that of the SUV_max alone.

Computed tomography-based radiomics for predicting lymphovascular invasion in rectal cancer

PURPOSE

To develop and externally validate a computed tomography (CT)-based radiomics model for predicting lymphovascular invasion (LVI) before treatment in patients with rectal cancer (RC).

METHOD

This retrospective study enrolled 351 patients with RC from three hospitals between March 2018 and March 2021. These patients were assigned to one of the following three groups: training set (n = 239, from hospital 1), internal validation set (n = 60, from hospital 1), and external validation set (n = 52, from hospitals 2 and 3). Large amounts of radiomics features were extracted from the intratumoral and peritumoral regions in the portal venous phase contrast-enhanced CT images. The score of radiomics features (Rad-score) was calculated by performing logistic regression analysis following the L1-based method. A combined model (Rad-score + clinical factors) was developed in the training cohort and validated internally and externally. The models were compared using the area under the receiver operating characteristic curve (AUC).

RESULTS

Of the 351 patients, 106 (30.2%) had an LVI + tumor. Rad-score (comprised of 22 features) was significantly higher in the LVI + group than in the LVI- group (0.60 ± 0.17 vs. 0.42 ± 0.19, P = 0.001). The combined model obtained good predictive performance in the training cohort (AUC = 0.813 [95% CI: 0.758-0.861]), with robust results in internal and external validations (AUC = 0.843 [95% CI: 0.726-0.924] and 0.807 [95% CI: 0.674-0.903]).

CONCLUSIONS

The proposed combined model demonstrated the potential to predict LVI preoperatively in patients with RC.

Male Breast Cancer Review. A Rare Case of Pure DCIS: Imaging Protocol, Radiomics and Management

Ductal carcinoma in situ (DCIS) of male breast is a rare lesion, often associated with invasive carcinoma. When the in situ component is present in pure form, histological grade is usually low or intermediate. Imaging is difficult as gynaecomastia is often present and can mask underlying findings. We report a rare case of pure high-grade DCIS in a young male patient, with associated intraductal papilloma and atypical ductal hyperplasia. Digital breast tomosynthesis (DBT) showed an area of architectural distortion at the union of outer quadrants of the left breast without gynaecomastia. Triple assessment suggested performing a nipple-sparing mastectomy, which revealed the presence of a focal area of high-grade DCIS of 2 mm. DCIS, even of high grade, is difficult to detect with mammography and even more rare, especially when associated with other proliferative lesions. DBT with 2D synthetic reconstruction is useful as the imaging step of a triple assessment and it should be performed in both symptomatic and asymptomatic high-risk men to differentiate between malignant and benign lesions. We propose a diagnostic model to early detect breast cancer in men, optimizing resources according to efficiency, effectiveness and economy, and look forward to radiomics as a powerful tool to help radiologists.

A Novel Approach Using FDG-PET/CT-Based Radiomics to Assess Tumor Immune Phenotypes in Patients With Non-Small Cell Lung Cancer

PURPOSE

Tumor microenvironment immune types (TMITs) are closely related to the efficacy of immunotherapy. We aimed to assess the predictive ability of 18F-fluorodeoxyglucose positron emission tomography/computed tomography (18F-FDG PET/CT)-based radiomics of TMITs in treatment-naive patients with non-small cell lung cancer (NSCLC).

METHODS

A retrospective analysis was performed in 103 patients with NSCLC who underwent 18F-FDG PET/CT scans. The patients were randomly assigned into a training set (n = 71) and a validation set (n = 32). Tumor specimens were analyzed by immunohistochemistry for the expression of programmed death-ligand 1 (PD-L1), programmed death-1 (PD-1), and CD8+ tumor-infiltrating lymphocytes (TILs) and categorized into four TMITs according to their expression of PD-L1 and CD8+ TILs. LIFEx package was used to extract radiomic features. The optimal features were selected using the least absolute shrinkage and selection operator (LASSO) algorithm, and a radiomics signature score (rad-score) was developed. We constructed a combined model based on the clinical variables and radiomics signature and compared the predictive performance of models using receiver operating characteristic (ROC) curves.

RESULTS

Four radiomic features (GLRLM_LRHGE, GLZLM_SZE, SUVmax, NGLDM_Contrast) were selected to build the rad-score. The rad-score showed a significant ability to discriminate between TMITs in both sets (p < 0.001, p < 0.019), with an area under the ROC curve (AUC) of 0.800 [95% CI (0.688-0.885)] in the training set and that of 0.794 [95% CI (0.615-0.916)] in the validation set, while the AUC values of clinical variables were 0.738 and 0.699, respectively. When clinical variables and radiomics signature were combined, the complex model showed better performance in predicting TMIT-I tumors, with the AUC values increased to 0.838 [95% CI (0.731-0.914)] in the training set and 0.811 [95% CI (0.634-0.927)] in the validation set.

CONCLUSION

The FDG-PET/CT-based radiomic features showed good performance in predicting TMIT-I tumors in NSCLC, providing a promising approach for the choice of immunotherapy in a clinical setting.

Brain MRI Radiomics Analysis of School-Aged Children with Tetralogy of Fallot

Introduction

Radiomics could be potential imaging biomarkers by capturing and analyzing the features. Children and adolescents with CHD have worse neurodevelopmental and functional outcomes compared with their peers. Early diagnosis and intervention are the necessity to improve neurological outcomes in CHD patients.

Methods

School-aged TOF patients and their healthy peers were recruited for MRI and neurodevelopmental assessment. LASSO regression was used for dimension reduction. ROC curve graph showed the performance of the model.

Results

Six related features were finally selected for modeling. The final model AUC was 0.750. The radiomics features can be potential significant predictors for neurodevelopmental diagnoses.

Conclusion

The radiomics on the conventional MRI can help predict the neurodevelopment of school-aged children and provide parents with rehabilitation advice as early as possible.

Comparison of Conventional and Radiomic Features between 18F-FBPA PET/CT and PET/MR

Boron-10-containing positron emission tomography (PET) radio-tracer, 18F-FBPA, has been used to evaluate the feasibility and treatment outcomes of Boron neutron capture therapy (BNCT). The clinical use of PET/MR is increasing and reveals its benefit in certain applications. However, the PET/CT is still the most widely used modality for daily PET practice due to its high quantitative accuracy and relatively low cost. Considering the different attenuation correction maps between PET/CT and PET/MR, comparison of derived image features from these two modalities is critical to identify quantitative imaging biomarkers for diagnosis and prognosis. This study aimed to investigate the comparability of image features extracted from 18F-FBPA PET/CT and PET/MR. A total of 15 patients with malignant brain tumor who underwent 18F-FBPA examinations using both PET/CT and PET/MR on the same day were retrospectively analyzed. Overall, four conventional imaging characteristics and 449 radiomic features were calculated from PET/CT and PET/MR, respectively. A linear regression model and intraclass correlation coefficient (ICC) were estimated to evaluate the comparability of derived features between two modalities. Features were classified into strong, moderate, and weak comparability based on coefficient of determination (r2) and ICC. All of the conventional features, 81.2% of histogram, 37.5% of geometry, 51.5% of texture, and 25% of wavelet-based features, showed strong comparability between PET/CT and PET/MR. With regard to the wavelet filtering, radiomic features without filtering (61.2%) or with low-pass filtering (59.2%) along three axes produced strong comparability between the two modalities. However, only 8.2% of the features with high-pass filtering showed strong comparability. The linear regression models were provided for the features with strong and moderate consensus to interchange the quantitative features between the PET/CT and the PET/MR. All of the conventional and 71% of the radiomic (mostly histogram and texture) features were sufficiently stable and could be interchanged between 18F-FBPA PET with different hybrid modalities using the proposed equations. Our findings suggested that the image features high interchangeability may facilitate future studies in comparing PET/CT and PET/MR.

Radiogenomics Map Reveals the Landscape of m6A Methylation Modification Pattern in Bladder Cancer

We aimed to develop a noninvasive radiomics approach to reveal the m6A methylation status and predict survival outcomes and therapeutic responses in patients. A total of 25 m6A regulators were selected for further analysis, we confirmed that expression level and genomic mutations rate of m6A regulators were significantly different between cancer and normal tissues. Besides, we constructed methylation modification models and explored the immune infiltration and biological pathway alteration among different models. The m6A subtypes identified in this study can effectively predict the clinical outcome of bladder cancer (including m6AClusters, geneClusters, and m6Ascore models). In addition, we observed that immune response markers such as PD1 and CTLA4 were significantly corelated with the m6Ascore. Subsequently, a total of 98 obtained digital images were processed to capture the image signature and construct image prediction models based on the m6Ascore classification using a radiomics algorithm. We constructed seven signature radiogenomics models to reveal the m6A methylation status, and the model achieved an area under curve (AUC) degree of 0.887 and 0.762 for the training and test datasets, respectively. The presented radiogenomics models, a noninvasive prediction approach that combined the radiomics signatures and genomics characteristics, displayed satisfactory effective performance for predicting survival outcomes and therapeutic responses of patients. In the future, more interdisciplinary fields concerning the combination of medicine and electronics remains to be explored.

Application of dual-stream 3D convolutional neural network based on 18F-FDG PET/CT in distinguishing benign and invasive adenocarcinoma in ground-glass lung nodules

Purpose

This work aims to train, validate, and test a dual-stream three-dimensional convolutional neural network (3D-CNN) based on fluorine 18 (¹⁸F)-fluorodeoxyglucose (FDG) PET/CT to distinguish benign lesions and invasive adenocarcinoma (IAC) in ground-glass nodules (GGNs).

Methods

We retrospectively analyzed patients with suspicious GGNs who underwent ¹⁸F-FDG PET/CT in our hospital from November 2011 to November 2020. The patients with benign lesions or IAC were selected for this study. According to the ratio of 7:3, the data were randomly divided into training data and testing data. Partial image feature extraction software was used to segment PET and CT images, and the training data after using the data augmentation were used for the training and validation (fivefold cross-validation) of the three CNNs (PET, CT, and PET/CT networks).

Results

A total of 23 benign nodules and 92 IAC nodules from 106 patients were included in this study. In the training set, the performance of PET network (accuracy, sensitivity, and specificity of 0.92 ± 0.02, 0.97 ± 0.03, and 0.76 ± 0.15) was better than the CT network (accuracy, sensitivity, and specificity of 0.84 ± 0.03, 0.90 ± 0.07, and 0.62 ± 0.16) (especially accuracy was significant, P-value was 0.001); in the testing set, the performance of both networks declined. However, the accuracy and sensitivity of PET network were still higher than that of CT network (0.76 vs. 0.67; 0.85 vs. 0.70). For dual-stream PET/CT network, its performance was almost the same as PET network in the training set (P-value was 0.372–1.000), while in the testing set, although its performance decreased, the accuracy and sensitivity (0.85 and 0.96) were still higher than both CT and PET networks. Moreover, the accuracy of PET/CT network was higher than two nuclear medicine physicians [physician 1 (3-year experience): 0.70 and physician 2 (10-year experience): 0.73].

Conclusion

The 3D-CNN based on ¹⁸F-FDG PET/CT can be used to distinguish benign lesions and IAC in GGNs, and the performance is better when both CT and PET images are used together.

Supplementary Information

The online version contains supplementary material available at 10.1186/s40658-021-00423-1.

Robustness and performance of radiomic features in diagnosing cystic renal masses

PURPOSE

We study the inter-reader variability in manual delineation of cystic renal masses (CRMs) presented in computerized tomography (CT) images and its effect on the classification performance of a machine learning algorithm in distinguishing benign from potentially malignant CRMs. In addition, we assessed whether the inclusion of higher-order robust radiomic features improves the classification performance over the use of first-order features.

METHODS

230 CRMs were independently delineated by two radiologists. Through a combination of random fluctuations, dilation, and erosion operations over the original region of interests (ROIs), we generated four additional sets of synthetic ROIs to capture the inter-reader variability realistically, as confirmed by dice coefficient measurements and visual assessment. We then identified the robust features based on the intra-class coefficient (ICC > 0.85) across these datasets. We applied a tenfold stratified cross-validation (CV) to train and test the performance of the random forest model for the classification of CRMs into benign and potentially malignant.

RESULTS

The mean area under the curve (AUC), sensitivity, specificity, positive predictive value, and negative predictive value were 0.87, 0.82, 0.90, 0.85, and 0.93, respectively. With the usage of first-order features alone, the corresponding values were nearly identical.

CONCLUSION

AUC ranged for the robust and uncorrelated features from 0.83 ± 0.09 to 0.93 ± 0.04 and for the first-order features from 0.84 ± 0.09 to 0.91 ± 0.04. Our study indicates that the first-order features alone are sufficient for the classification of CRMs, and that inclusion of higher-order features does not necessarily improve performance.

Performance of automatic machine learning versus radiologists in the evaluation of endometrium on computed tomography

PURPOSE

In this study, we developed radiomic models that utilize a combination of imaging features and clinical variables to distinguish endometrial cancer (EC) from normal endometrium on routine computed tomography (CT).

METHODS

A total of 926 patients consisting of 416 endometrial cancer (EC) and 510 normal endometrium were included. The CT images of these patients were segmented manually, and divided into training, validation, testing and external testing sets. Non-texture and texture features of these images with endometrium or uterus as region of interest were extracted. The clinical feature "age" was also included in the feature set. Feature selection and machine learning classifier were applied to normalized feature set. This manual optimized combination was then compared with the best pipeline exported by Tree-Based Pipeline Optimization Tool (TPOT) on testing and external testing set. The performances of these machine learning pipelines were compared to that of radiologists.

RESULTS

The manual expert optimized pipeline using the "reliefF" feature selection method and "Bagging" classifier on the external testing set achieved a test ROC AUC of 0.73, accuracy of 0.73 (95% CI 0.62-0.82), sensitivity of 0.64 (95% CI 0.45-0.79), and specificity of 0.78 (95% CI 0.65-0.87), while TPOT achieved a test ROC AUC of 0.79, accuracy of 0.80 (95% CI 0.70-0.87), sensitivity of 0.61 (95% CI 0.43-0.77), and specificity of 0.90 (95% CI 0.78-0.96). When compared to average radiologist performance, the TPOT achieved higher test accuracy (0.80 vs. 0.49, p < 0.001) and specificity (0.90 vs. 0.51, p < 0.001), with comparable sensitivity (0.61 vs. 0.46, p = 0.130).

CONCLUSION

Our results demonstrate that automatic machine learning can distinguish EC from normal endometrium on routine CT imaging with higher accuracy and specificity than radiologists.

Baseline 18F-FDG PET textural features as predictors of response to chemotherapy in diffuse large B-cell lymphoma

PURPOSE

We sought to investigate the performance of radiomics analysis on baseline 18F-FDG PET/CT for predicting response to first-line chemotherapy in diffuse large B-cell lymphoma (DLBCL).

MATERIAL AND METHODS

Forty-five patients who received first-line rituximab, cyclophosphamide, doxorubicin, vincristine and prednisone (R-CHOP) chemotherapy for DLBCL were included in the study. Radiomics features and standard uptake value (SUV)-based measurements were extracted from baseline PET images for a total of 147 lesions. The selection of the most relevant features was made using the recursive feature elimination algorithm. A machine-learning model was trained using the logistic regression classifier with cross-validation to predict treatment response. The independent predictors of incomplete response were evaluated with multivariable regression analysis.

RESULTS

A total of 14 textural features were selected by the recursive elimination algorithm, achieving a feature-to-lesion ratio of 1:10. The accuracy and area under the receiver operating characteristic curve of the model for predicting incomplete response were 0.87 and 0.81, respectively. Multivariable analysis revealed that SUVmax and gray level co-occurrence matrix dissimilarity were independent predictors of lesions with incomplete response to first-line R-CHOP chemotherapy.

CONCLUSION

Increased textural heterogeneity in baseline PET images was found to be associated with incomplete response in DLBCL.

CT-Based Peritumoral and Intratumoral Radiomics as Pretreatment Predictors of Atypical Responses to Immune Checkpoint Inhibitor Across Tumor Types: A Preliminary Multicenter Study

OBJECTIVE

To develop and validate a new strategy based on radiomics features extracted from intra- and peritumoral regions on CT images for the prediction of atypical responses to the immune checkpoint inhibitor (ICI) in cancer patients.

METHODS

In total, 135 patients derived from five hospitals with pathologically confirmed malignancies receiving ICI were included in this retrospective study. Atypical responses including pseudoprogression (PsP) and hyperprogression disease (HPD) were identified as their definitions. A subgroup of standard progression disease (sPD) in 2018 was also involved in this study. Based on pretreatment CT imaging, a total of 107 features were extracted from intra- and peri-tumoral regions, respectively. The least absolute shrinkage and selection operator (Lasso) algorithm was used for feature selection, and multivariate logistic analysis was used to develop radiomics signature (RS). Finally, a total of nine RSs, derived from intra-tumoral, peri-tumoral, and combination of both regions, were built respectively to distinguish PsP vs. HPD, PsP vs. sPD, and HPD vs. sPD. The performance of the RSs was evaluated with discrimination, calibration, and clinical usefulness.

RESULTS

No significant difference was found when compared in terms of clinical characteristics of PsP, HPD, and sPD. RS based on combined regions outperformed those from either intra-tumoral or peri-tumoral alone, yielding an AUC (accuracy) of 0.834 (0.827) for PsP vs. HPD, 0.923 (0.868) for PsP vs. sPD, and 0.959 (0.894) for HPD vs. sPD in the training datasets, and 0.835 (0.794) for PsP vs. HPD, 0.919 (0.867) for PsP vs. sPD, and 0.933 (0.842) for HPD vs. sPD in the testing datasets. The combined RS showed good fitness (Hosmer-Lemeshow test p > 0.05) and provided more net benefit than the treat-none or treat-all scheme by decision curve analysis in both training and testing datasets.

CONCLUSION

Pretreatment radiomics are helpful to predict atypical responses to ICI across tumor types. The combined RS outperformed those from either intra- or peri-tumoral alone which may provide a more comprehensive characterization of atypical responses to ICI.

Diagnosis of Ovarian Neoplasms Using Nomogram in Combination With Ultrasound Image-Based Radiomics Signature and Clinical Factors

Objectives: To establish and validate a nomogram integrating radiomics signatures from ultrasound and clinical factors to discriminate between benign, borderline, and malignant serous ovarian tumors. Materials and methods: In this study, a total of 279 pathology-confirmed serous ovarian tumors collected from 265 patients between March 2013 and December 2016 were used. The training cohort was generated by randomly selecting 70% of each of the three types (benign, borderline, and malignant) of tumors, while the remaining 30% was included in the validation cohort. From the transabdominal ultrasound scanning of ovarian tumors, the radiomics features were extracted, and a score was calculated. The ability of radiomics to differentiate between the grades of ovarian tumors was tested by comparing benign vs borderline and malignant (task 1) and borderline vs malignant (task 2). These results were compared with the diagnostic performance and subjective assessment by junior and senior sonographers. Finally, a clinical-feature alone model and a combined clinical-radiomics (CCR) model were built using predictive nomograms for the two tasks. Receiver operating characteristic (ROC) analysis, calibration curve, and decision curve analysis (DCA) were performed to evaluate the model performance. Results: The US-based radiomics models performed satisfactorily in both the tasks, showing especially higher accuracy in the second task by successfully discriminating borderline and malignant ovarian serous tumors compared to the evaluations by senior sonographers (AUC = 0.789 for seniors and 0.877 for radiomics models in task one; AUC = 0.612 for senior and 0.839 for radiomics model in task 2). We showed that the CCR model, comprising CA125 level, lesion location, ascites, and radiomics signatures, performed the best (AUC = 0.937, 95%CI 0.905-0.969 in task 1, AUC = 0.924, 95%CI 0.876-0.971 in task 2) in the training as well as in the validation cohorts (AUC = 0.914, 95%CI 0.851-0.976 in task 1, AUC = 0.890, 95%CI 0.794-0.987 in task 2). The calibration curve and DCA analysis of the CCR model more accurately predicted the classification of the tumors than the clinical features alone. Conclusion: This study integrates novel radiomics signatures from ultrasound and clinical factors to create a nomogram to provide preoperative diagnostic information for differentiating between benign, borderline, and malignant ovarian serous tumors, thereby reducing unnecessary and risky biopsies and surgeries.

Current Advances and Challenges in Radiomics of Brain Tumors

Imaging diagnosis is crucial for early detection and monitoring of brain tumors. Radiomics enable the extraction of a large mass of quantitative features from complex clinical imaging arrays, and then transform them into high-dimensional data which can subsequently be mined to find their relevance with the tumor's histological features, which reflect underlying genetic mutations and malignancy, along with grade, progression, therapeutic effect, or even overall survival (OS). Compared to traditional brain imaging, radiomics provides quantitative information linked to meaningful biologic characteristics and application of deep learning which sheds light on the full automation of imaging diagnosis. Recent studies have shown that radiomics' application is broad in identifying primary tumor, differential diagnosis, grading, evaluation of mutation status and aggression, prediction of treatment response and recurrence in pituitary tumors, gliomas, and brain metastases. In this descriptive review, besides establishing a general understanding among protocols, results, and clinical significance of these studies, we further discuss the current limitations along with future development of radiomics.

Radiomics in PET Imaging:: A Practical Guide for Newcomers

Radiomics has undergone considerable development in recent years. In PET imaging, very promising results concerning the ability of handcrafted features to predict the biological characteristics of lesions and to assess patient prognosis or response to treatment have been reported in the literature. This article presents a checklist for designing a reliable radiomic study, gives an overview of the steps of the pipeline, and outlines approaches for data harmonization. Tips are provided for critical reading of the content of articles. The advantages and limitations of handcrafted radiomics compared with deep-learning approaches for the characterization of PET images are also discussed.

Prediction of Clinical Outcome for High-Intensity Focused Ultrasound Ablation of Uterine Leiomyomas Using Multiparametric MRI Radiomics-Based Machine Leaning Model

Objectives

This study sought to develop a multiparametric MRI radiomics-based machine learning model for the preoperative prediction of clinical success for high-intensity-focused ultrasound (HIFU) ablation of uterine leiomyomas.

Methods

One hundred and thirty patients who received HIFU ablation therapy for uterine leiomyomas were enrolled in this retrospective study. Radiomics features were extracted from T2-weighted (T2WI) image and ADC map derived from diffusion-weighted imaging (DWI). Three feature selection algorithms including least absolute shrinkage and selection operator (LASSO), recursive feature elimination (RFE), and ReliefF algorithm were used to select radiomics features, respectively, which were fed into four machine learning classifiers including k-nearest neighbors (KNN), logistic regression (LR), random forest (RF), and support vector machine (SVM) for the construction of outcome prediction models before HIFU treatment. The performance, predication ability, and clinical usefulness of these models were verified and evaluated using receiver operating characteristics (ROC), calibration, and decision curve analyses.

Results

The radiomics analysis provided an effective preoperative prediction for HIFU ablation of uterine leiomyomas. Using SVM with ReliefF algorithm, the multiparametric MRI radiomics model showed the favorable performance with average accuracy of 0.849, sensitivity of 0.814, specificity of 0.896, positive predictive value (PPV) of 0.903, negative predictive value (NPV) of 0.823, and the area under the ROC curve (AUC) of 0.887 (95% CI = 0.848-0.939) in fivefold cross-validation, followed by RF with ReliefF. Calibration and decision curve analyses confirmed the potential of model in predication ability and clinical usefulness.

Conclusions

The radiomics-based machine learning model can predict preoperatively HIFU ablation response for the patients with uterine leiomyomas and contribute to determining individual treatment strategies.

Artificial Intelligence in Thyroid Field-A Comprehensive Review

Simple Summary

The incidence of thyroid pathologies has been increasing worldwide. Historically, the detection of thyroid neoplasms relies on medical imaging analysis, depending mainly on the experience of clinicians. The advent of artificial intelligence (AI) techniques led to a remarkable progress in image-recognition tasks. AI represents a powerful tool that may facilitate understanding of thyroid pathologies, but actually, the diagnostic accuracy is uncertain. This article aims to provide an overview of the basic aspects, limitations and open issues of the AI methods applied to thyroid images. Medical experts should be familiar with the workflow of AI techniques in order to avoid misleading outcomes.

Abstract

Artificial intelligence (AI) uses mathematical algorithms to perform tasks that require human cognitive abilities. AI-based methodologies, e.g., machine learning and deep learning, as well as the recently developed research field of radiomics have noticeable potential to transform medical diagnostics. AI-based techniques applied to medical imaging allow to detect biological abnormalities, to diagnostic neoplasms or to predict the response to treatment. Nonetheless, the diagnostic accuracy of these methods is still a matter of debate. In this article, we first illustrate the key concepts and workflow characteristics of machine learning, deep learning and radiomics. We outline considerations regarding data input requirements, differences among these methodologies and their limitations. Subsequently, a concise overview is presented regarding the application of AI methods to the evaluation of thyroid images. We developed a critical discussion concerning limits and open challenges that should be addressed before the translation of AI techniques to the broad clinical use. Clarification of the pitfalls of AI-based techniques results crucial in order to ensure the optimal application for each patient.

Radiomic Features on Multiparametric MRI for Preoperative Evaluation of Pituitary Macroadenomas Consistency: Preliminary Findings

BACKGROUND

Preoperative assessment of the consistency of pituitary macroadenomas (PMA) might be needed for surgical planning.

PURPOSE

To investigate the diagnostic performance of radiomics models based on multiparametric magnetic resonance imaging (mpMRI) for preoperatively evaluating the tumor consistency of PMA.

STUDY TYPE

Retrospective.

POPULATION

One hundred and fifty-six PMA patients (soft consistency, N = 104 vs. hard consistency, N = 52), divided into training (N = 108) and test (N = 48) cohorts. The tumor consistency was determined on surgical findings.

FIELD STRENGTH/SEQUENCE

T1-weighted imaging (T1WI), contrast-enhanced T1WI (T1CE), and T2-weighted imaging (T2WI) using spin-echo sequences with a 3.0-T scanner.

ASSESSMENT

An automated three-dimensional (3D) segmentation was performed to generate the volume of interest (VOI) on T2WI, then T1WI/T1CE were coregistered to T2WI. A total of 388 radiomic features were extracted on each VOI of mpMRI. The top-discriminative features were identified using the minimum-redundancy maximum-relevance method and 0.632+ bootstrapping. The radiomics models based on each sequence and their combinations were established via the random forest (RF) and support vector machine (SVM), and independently evaluated for their ability in distinguishing PMA consistency.

STATISTICAL TESTS

Mann-Whitney U-test and Chi-square test were used for comparison analysis. The area under the receiver operating characteristic curve (AUC), accuracy (ACC), sensitivity (SEN), specificity (SPE), and relative standard deviation (RSD) were calculated to evaluate each model's performance. ACC with P-value<0.05 was considered statistically significant.

RESULTS

Eleven mpMRI-based features exhibited statistically significant differences between soft and hard PMA in the training cohort. The radiomics model built on combined T1WI/T1CE/T2WI demonstrated the best performance among all the radiomics models with an AUC of 0.90 (95% confidence interval [CI]: 0.87-0.92), ACC of 0.87 (CI: 0.84-0.89), SEN of 0.83 (CI: 0.81-0.85), and SPE of 0.87 (CI: 0.85-0.99) in the test cohort.

DATA CONCLUSION

Radiomic features based on mpMRI have good performance in the presurgical evaluation of PMA consistency.

LEVEL OF EVIDENCE

3 TECHNICAL EFFICACY: Stage 2.

A Radiomics Signature-Based Nomogram to Predict the Progression-Free Survival of Patients With Hepatocellular Carcinoma After Transcatheter Arterial Chemoembolization Plus Radiofrequency Ablation

Objective: The study aims to establish an magnetic resonance imaging radiomics signature-based nomogram for predicting the progression-free survival of intermediate and advanced hepatocellular carcinoma (HCC) patients treated with transcatheter arterial chemoembolization (TACE) plus radiofrequency ablation

Materials and Methods: A total of 113 intermediate and advanced HCC patients treated with TACE and RFA were eligible for this study. Patients were classified into a training cohort (n = 78 cases) and a validation cohort (n = 35 cases). Radiomics features were extracted from contrast-enhanced T1W images by analysis kit software. Dimension reduction was conducted to select optimal features using the least absolute shrinkage and selection operator (LASSO). A rad-score was calculated and used to classify the patients into high-risk and low-risk groups and further integrated into multivariate Cox analysis. Two prediction models based on radiomics signature combined with or without clinical factors and a clinical model based on clinical factors were developed. A nomogram comcined radiomics signature and clinical factors were established and the concordance index (C-index) was used for measuring discrimination ability of the model, calibration curve was used for measuring calibration ability, and decision curve and clinical impact curve are used for measuring clinical utility.

Results: Eight radiomics features were selected by LASSO, and the cut-off of the Rad-score was 1.62. The C-index of the radiomics signature for PFS was 0.646 (95%: 0.582–0.71) in the training cohort and 0.669 (95% CI:0.572–0.766) in validation cohort. The median PFS of the low-risk group [30.4 (95% CI: 19.41–41.38)] months was higher than that of the high-risk group [8.1 (95% CI: 4.41–11.79)] months in the training cohort (log rank test, z = 16.58, p < 0.001) and was verified in the validation cohort. Multivariate Cox analysis showed that BCLC stage [hazard ratio (HR): 2.52, 95% CI: 1.42–4.47, p = 0.002], AFP level (HR: 2.01, 95% CI: 1.01–3.99 p = 0.046), time interval (HR: 0.48, 95% CI: 0.26–0.87, p = 0.016) and radiomics signature (HR 2.98, 95% CI: 1.60–5.51, p = 0.001) were independent prognostic factors of PFS in the training cohort. The C-index of the combined model in the training cohort was higher than that of clinical model for PFS prediction [0.722 (95% CI: 0.657–0.786) vs. 0.669 (95% CI: 0.657–0.786), p＜0.001]. Similarly, The C-index of the combined model in the validation cohort, was higher than that of clinical model [0.821 (95% CI: 0.726–0.915) vs. 0.76 (95% CI: 0.667–0.851), p = 0.004]. The calibration curve, decision curve and clinical impact curve showed that the nomogram can be used to accurately predict the PFS of patients.

Conclusion: The radiomics signature was a prognostic risk factor, and a nomogram combined radiomics and clinical factors acts as a new strategy for predicted the PFS of intermediate and advanced HCC treated with TACE plus RFA.

CT-Based Radiomics Nomogram for Prediction of Progression-Free Survival in Locoregionally Advanced Nasopharyngeal Carcinoma

Purpose

We aimed to construct of a nomogram to predict progression-free survival (PFS) in locoregionally advanced nasopharyngeal carcinoma (LA-NPC) with risk stratification using computed tomography (CT) radiomics features and clinical factors.

Patients and Methods

A total of 311 patients diagnosed with LA-NPC (stage III–IVa) at our hospital between 2010 and 2014 were included. The region of interest (ROI) of the primary nasopharyngeal mass was manually outlined. Independent sample t-test and LASSO-logistic regression were used for selecting the most predictive radiomics features of PFS, and to generate a radiomics signature. A nomogram was built with clinical factors and radiomics features, and the risk stratification model was tested accordingly.

Results

In total, 20 radiomics features most associated with prognosis were selected. The radiomics nomogram, which integrated the radiomics signature and significant clinical factors, showed excellent performance in predicting PFS, with C-index of 0.873 (95% CI: 0.803~0.943), which was better than that of the clinical nomogram (C-index, 0.729, 95% CI: 0.620~0.838) as well as of the TNM staging system (C-index, 0.689, 95% CI: 0.592–0.787) in validation cohort. The calibration curves and the decision curve analysis (DCA) plot obtained suggested satisfying accuracy and clinical utility of the model. The risk stratification tool was able to predict differences in prognosis of patients in different risk categories (p<0.001).

Conclusion

CT-based radiomics features, an in particular, radiomics nomograms, have the potential to become an accurate and reliable tool for assisting with prognosis prediction of LA-NPC.

Magnetic resonance imaging-based radiogenomics analysis for predicting prognosis and gene expression profile in advanced nasopharyngeal carcinoma

BACKGROUND

To establish a radiomics nomogram for survival prediction and determine if genomic data were related to radiomics signature in advanced nasopharyngeal carcinoma (NPC).

METHODS

Radiomics features were extracted from contrast-enhanced T1-weighted images (CE-T1WI) in 316 patients. A progression-free survival (PFS) nomogram was developed and validated by the combination of the radiomics signature and clinicopathologic factors. Whole transcriptomics sequencing was performed in pretreatment tumor samples; correlation of gene expression and radiomics signature was further investigated.

RESULTS

A 24-feature-combined radiomics signature was highly correlated with PFS; its integration with clinical predictors showed good prediction performance in the training and the validation cohort (C-index: 0.80 and 0.73). A significant correlation was observed between certain gene expression and Rad-score, especially the mRNA expression of CDKL2, PLIN5, and SPAG1.

CONCLUSION

As a noninvasive method, the MRI-based radiomics signature might enable the pretreatment prediction of prognosis and gene expressions profile in advanced NPC.

NSCLC Biomarkers to Predict Response to Immunotherapy with Checkpoint Inhibitors (ICI): From the Cells to In Vivo Images

Lung cancer remains the leading cause of cancer-related death, and it is usually diagnosed in advanced stages (stage III or IV). Recently, the availability of targeted strategies and of immunotherapy with checkpoint inhibitors (ICI) has favorably changed patient prognosis. Treatment outcome is closely related to tumor biology and interaction with the tumor immune microenvironment (TME). While the response in molecular targeted therapies relies on the presence of specific genetic alterations in tumor cells, accurate ICI biomarkers of response are lacking, and clinical outcome likely depends on multiple factors that are both host and tumor-related. This paper is an overview of the ongoing research on predictive factors both from in vitro/ex vivo analysis (ranging from conventional pathology to molecular biology) and in vivo analysis, where molecular imaging is showing an exponential growth and use due to technological advancements and to the new bioinformatics approaches applied to image analyses that allow the recovery of specific features in specific tumor subclones.

Multiparameter MRI-based radiomics for preoperative prediction of extramural venous invasion in rectal cancer

OBJECTIVES

To compare multiparameter MRI-based radiomics for preoperative prediction of extramural venous invasion (EMVI) in rectal cancer using different machine learning algorithms and to develop and validate the best diagnostic model.

METHODS

We retrospectively analyzed 317 patients with rectal cancer. Of these, 114 were EMVI positive and 203 were EMVI negative. Radiomics features were extracted from T₂-weighted imaging, T₁-weighted imaging, diffusion-weighted imaging, and enhanced T₁-weighted imaging of rectal cancer, followed by the dimension reduction of the features. Logistic regression, support vector machine, Bayes, K-nearest neighbor, and random forests algorithms were trained to obtain the radiomics signatures. The area under the receiver operating characteristic curve (AUC) was used to evaluate the performance of each radiomics signature. The best radiomics signature was selected and combined with clinical and radiological characteristics to construct a joint model for predicting EMVI. Finally, the predictive performance of the joint model was assessed.

RESULTS

The Bayes-based radiomics signature performed well in both the training set and the test set, with the AUCs of 0.744 and 0.738, sensitivities of 0.754 and 0.728, and specificities of 0.887 and 0.918, respectively. The joint model performed best in both the training set and the test set, with the AUCs of 0.839 and 0.835, sensitivities of 0.633 and 0.714, and specificities of 0.901 and 0.885, respectively.

CONCLUSIONS

The joint model demonstrated the best diagnostic performance for the preoperative prediction of EMVI in patients with rectal cancer. Hence, it can be used as a key tool for clinical individualized EMVI prediction.

KEY POINTS

• Radiomics features from magnetic resonance imaging can be used to predict extramural venous invasion (EMVI) in rectal cancer. • Machine learning can improve the accuracy of predicting EMVI in rectal cancer. • Radiomics can serve as a noninvasive biomarker to monitor the status of EMVI.

Prediction of Genetic Alterations in Oncogenic Signaling Pathways in Squamous Cell Carcinoma of the Head and Neck: Radiogenomic Analysis Based on Computed Tomography Images

OBJECTIVE

This study investigated the role of radiomics in evaluating the alterations of oncogenic signaling pathways in head and neck cancer.

METHODS

Radiomics features were extracted from 106 enhanced computed tomography images with head and neck squamous cell carcinoma. Support vector machine-recursive feature elimination was used for feature selection. Support vector machine algorithm was used to develop radiomics scores to predict genetic alterations in oncogenic signaling pathways. The performance was evaluated by the area under the curve (AUC) of the receiver operating characteristic curve.

RESULTS

The alterations of the Cell Cycle, HIPPO, NOTCH, PI3K, RTK RAS, and TP53 signaling pathways were predicted by radiomics scores. The AUC values of the training cohort were 0.94, 0.91, 0.94, 0.93, 0.87, and 0.93, respectively. The AUC values of the validation cohort were all greater than 0.7.

CONCLUSIONS

Radiogenomics is a new method for noninvasive acquisition of tumor molecular information at the genetic level.