Comparison of MRI and CT-based radiomics for preoperative prediction of lymph node metastasis in pancreatic ductal adenocarcinoma.

BACKGROUND: The preoperative prediction of lymph node metastasis (LNM) in pancreatic ductal adenocarcinoma (PDAC) is essential in prognosis and treatment strategy formulation. PURPOSE: To compare the performance of computed tomography (CT) and magnetic resonance imaging (MRI) radiomics models for the preoperative prediction of LNM in PDAC. MATERIAL AND METHODS: In total, 160 consecutive patients with PDAC were retrospectively included, who were divided into the training and validation sets (ratio of 8:2). Two radiologists evaluated LNM basing on morphological abnormalities. Radiomics features were extracted from T2-weighted imaging, T1-weighted imaging, and multiphase contrast enhanced MRI and multiphase CT, respectively. Overall, 1184 radiomics features were extracted from each volume of interest drawn. Only features with an intraclass correlation coefficient ≥0.75 were included. Three sequential feature selection steps-variance threshold, variance thresholding and least absolute shrinkage selection operator-were repeated 20 times with fivefold cross-validation in the training set. Two radiomics models based on multiphase CT and multiparametric MRI were built with the five most frequent features. Model performance was evaluated using the area under the curve (AUC) values. RESULTS: Multiparametric MRI radiomics model achieved improved AUCs (0.791 and 0.786 in the training and validation sets, respectively) than that of the CT radiomics model (0.672 and 0.655 in the training and validation sets, respectively) and of the radiologists' assessment (0.600-0.613 and 0.560-0.587 in the training and validation sets, respectively). CONCLUSION: Multiparametric MRI radiomics model may serve as a potential tool for preoperatively evaluating LNM in PDAC and had superior predictive performance to multiphase CT-based model and radiologists' assessment.

Application of Magnetic Resonance Imaging in Neoadjuvant Treatment of Pancreatic Ductal Adenocarcinoma.

Pancreatic ductal adenocarcinoma (PDAC) is one of the deadliest malignant tumors of the human digestive system. Due to its insidious onset, many patients have already lost the opportunity for radical resection upon tumor diagnosis. In recent years, neoadjuvant treatment for patients with borderline resectable PDAC has been recommended by multiple guidelines to increase the resection rate of radical surgery and improve the postoperative survival. However, further developments are required to accurately assess the tumor response to neoadjuvant therapy and to select the population suitable for such treatment. Reductions in drug toxicity and the number of neoadjuvant cycles are also critical. At present, the clinical evaluation of neoadjuvant treatment is mainly based on several serological and imaging indicators; however, the unique characteristics of PDAC and the insufficient sensitivity and specificity of the markers render this system ineffective. The imaging evaluation system, magnetic resonance imaging (MRI), has its own unique imaging advantages compared with computed tomography (CT) and other imaging examinations. One key advantage is the ability to reflect the changes more rapidly in tumor tissue components, such as the degree of fibrosis, microvessel density, and tissue hypoxia. It can also perform multiparameter quantitative analysis of tumor tissue and changes, attributing to its increasingly important role in imaging evaluation, and potentially the evaluation of neoadjuvant treatment of pancreatic cancer, as several current articles have studied. At the same time, owing to the complexity of MRI and some of its limitations, its wider application is limited. Compared with CT imaging, few relevant studies have been conducted. In this review article, we will investigate and summarize the advantages, limitations, and future development of MRI in the evaluation of neoadjuvant treatment of PDAC. EVIDENCE LEVEL: 3 TECHNICAL EFFICACY: Stage 2.

A Deep Learning Approach for MRI in the Diagnosis of Labral Injuries of the Hip Joint.

BACKGROUND: The diagnosis of labral injury on MRI is time-consuming and potential for incorrect diagnoses. PURPOSE: To explore the feasibility of applying deep learning to diagnose and classify labral injuries with MRI. STUDY TYPE: Retrospective. POPULATION: A total of 1016 patients were divided into normal (n = 168, class 0) and abnormal labrum (n = 848) groups. The abnormal group consisted of n = 111 with class 1 (degeneration), n = 437 with class 2 (partial or complete tear), and n = 300 with unclassified injury. Patients were randomly divided into training, validation, and test cohort according to the ratio of 55%:15%:30%. FIELD STRENGTH/SEQUENCE: Fat-saturation proton density-weighted fast spin-echo sequence at 3.0 T. ASSESSMENT: Convolutional neural network-6 (CNN-6) was used to extract, discriminate, and detect oblique coronal (OCOR) and oblique sagittal (OSAG) images. Mask R-CNN was used for segmentation. LeNet-5 was used to diagnose and classify labral injuries. The weighting method combined the models of OCOR and OSAG. The output-input connection was used to correlate the whole diagnosis/classification system. Four radiologists performed subjective diagnoses to obtain the diagnosis results. STATISTICAL TESTS: CNN-6 and LeNet-5 were evaluated by area under the receiver operating characteristic (ROC) curve and related parameters. The mean average precision (MAP) evaluated the Mask R-CNN. McNemar's test was used to compare the radiologists and models. A P value < 0.05 was considered statistically significant. RESULTS: The area under the curve (AUC) of CNN-6 was 0.99 for extraction, discrimination, and detection. MAP values of Mask R-CNN for OCOR and OSAG image segmentation were 0.96 and 0.99. The accuracies of LeNet-5 in the diagnosis and classification were 0.94/0.94 (OCOR) and 0.92/0.91 (OSAG), respectively. The accuracy of the weighted models in the diagnosis and classification were 0.94 and 0.97, respectively. The accuracies of radiologists in the diagnosis and classification of labrum injuries ranged from 0.85 to 0.92 and 0.78 to 0.94, respectively. DATA CONCLUSION: Deep learning can assist radiologists in diagnosing and classifying labrum injuries. EVIDENCE LEVEL: 3 TECHNICAL EFFICACY: Stage 2.

Vertebral MRI-based radiomics model to differentiate multiple myeloma from metastases: influence of features number on logistic regression model performance.

OBJECTIVES: This study aimed to use the most frequent features to establish a vertebral MRI-based radiomics model that could differentiate multiple myeloma (MM) from metastases and compare the model performance with different features number. METHODS: We retrospectively analyzed conventional MRI (T1WI and fat-suppression T2WI) of 103 MM patients and 138 patients with metastases. The feature selection process included four steps. The first three steps defined as conventional feature selection (CFS), carried out 50 times (ten times with 5-fold cross-validation), included variance threshold, SelectKBest, and least absolute shrinkage and selection operator. The most frequent fixed features were selected for modeling during the last step. The number of events per independent variable (EPV) is the number of patients in a smaller subgroup divided by the number of radiomics features considered in developing the prediction model. The EPV values considered were 5, 10, 15, and 20. Therefore, we constructed four models using the top 16, 8, 6, and 4 most frequent features, respectively. The models constructed with features selected by CFS were also compared. RESULTS: The AUCs of 20EPV-Model, 15EPV-Model, and CSF-Model (AUC = 0.71, 0.81, and 0.78) were poor than 10EPV-Model (AUC = 0.84, p < 0.001). The AUC of 10EPV-Model was comparable with 5EPV-Model (AUC = 0.85, p = 0.480). CONCLUSIONS: The radiomics model constructed with an appropriate small number of the most frequent features could well distinguish metastases from MM based on conventional vertebral MRI. Based on our results, we recommend following the 10 EPV as the rule of thumb for feature selection. KEY POINTS: • The developed radiomics model could distinguish metastases from multiple myeloma based on conventional vertebral MRI. • An accurate model based on just a handful of the most frequent features could be constructed by utilizing multiple feature reduction techniques. • An event per independent variable value of 10 is recommended as a rule of thumb for modeling feature selection.

Prediction of High-Risk Cytogenetic Status in Multiple Myeloma Based on Magnetic Resonance Imaging: Utility of Radiomics and Comparison of Machine Learning Methods.

BACKGROUND: Radiomics has shown promising results in the diagnosis, efficacy, and prognostic assessments of multiple myeloma (MM). However, little evidence exists on the utility of radiomics in predicting a high-risk cytogenetic (HRC) status in MM. PURPOSE: To develop and test a magnetic resonance imaging (MRI)-based radiomics model for predicting an HRC status in MM patients. STUDY TYPE: Retrospective. POPULATION: Eighty-nine MM patients (HRC [n: 37] and non-HRC [n: 52]). FIELD STRENGTH/SEQUENCE: A 3.0 T; fast spin-echo (FSE): T1-weighted image (T1WI) and fat-suppression T2WI (FS-T2WI). ASSESSMENT: Overall, 1409 radiomics features were extracted from each volume of interest drawn by radiologists. Three sequential feature selection steps-variance threshold, SelectKBest, and least absolute shrinkage selection operator-were repeated 10 times with 5-fold cross-validation. Radiomics models were constructed with the top three frequency features of T1 WI/T2 WI/two-sequence MRI (T1 WI and FS-T2 WI). Radiomics models, clinical data (age and visually assessed MRI pattern), or radiomics combined with clinical data were used with six classifiers to distinguish between HRC and non-HRC statuses. Six classifiers used were support vector machine, random forest, logistic regression (LR), decision tree, k-nearest neighbor, and XGBoost. Model performance was evaluated with area under the curve (AUC) values. STATISTICAL TESTS: Mann-Whitney U-test, Chi-squared test, Z test, and DeLong method. RESULTS: The LR classifier performed better than the other classifiers based on different data (AUC: 0.65-0.82; P < 0.05). The two-sequence MRI models performed better than the other data models using different classifiers (AUC: 0.68-0.82; P < 0.05). Thus, the LR two-sequence model yielded the best performance (AUC: 0.82 ± 0.02; sensitivity: 84.1%; specificity: 68.1%; accuracy: 74.7%; P < 0.05). CONCLUSION: The LR-based machine learning method appears superior to other classifier methods for assessing HRC in MM. Radiomics features based on two-sequence MRI showed good performance in differentiating HRC and non-HRC statuses in MM. EVIDENCE LEVEL: 3 TECHNICAL EFFICACY: Stage 2.

A preliminary study using spinal MRI-based radiomics to predict high-risk cytogenetic abnormalities in multiple myeloma.

OBJECTIVES: We aimed to investigate the feasibility of predicting high-risk cytogenetic abnormalities (HRCAs) in patients with multiple myeloma (MM) using a spinal MRI-based radiomics method. MATERIALS AND METHODS: In this retrospective study, we analyzed the radiomic features of 248 lesions (HRCA [n = 111] and non-HRCA [n = 137]) using T1WI, T2WI, and fat suppression T2WI. To construct the radiomics model, the top nine most frequent radiomic features were selected using logistic regression (LR) machine-learning processes. A combined LR model incorporating radiomic features and basic clinical characteristics (age and sex) was also built. Fivefold external cross-validation was performed, and a comparative analysis of 10 random fivefold cross-validation sets was used to verify result stability. Model performance was compared by plotting receiver operating characteristic curves and the area under the curve (AUC). RESULTS: Comparable AUC values were observed between the radiomics model and the combined model in validation cohorts (AUC: 0.863 vs. 0.870, respectively, p = 0.206). The radiomics model had an AUC of 0.863, with a sensitivity of 0.789, a specificity of 0.787, a positive predictive value of 0.753, a negative predictive value of 0.824, and an accuracy of 0.788 in the validation cohort, which were comparable with the performance in the training cohorts. CONCLUSIONS: Radiomic features of routine spinal MRI reflect differences between HRCAs and non-HRCAs in patients with MM. This MRI-based radiomics model might be a useful and independent tool to predict HRCAs in patients MM.

The diagnostic value of intravoxel incoherent motion diffusion-weighted imaging for distinguishing nonhypervascular pancreatic neuroendocrine tumors from pancreatic ductal adenocarcinomas.

PURPOSE: To primarily evaluate the diagnostic performance of the monoexponential and intravoxel incoherent motion (IVIM) diffusion weighted imaging (DWI) models for differentiating between nonhypervascular pancreatic neuroendocrine tumors (PNETs) and pancreatic ductal adenocarcinomas (PDACs). METHODS: 63 patients with PNETs (35 nonhypervascular PNETs and 28 hypervascular PNETs) and 164 patients with PDACs were retrospectively enrolled in the study and underwent multiple b-value DWI. Intraobserver and interobserver reliabilities of DWI parameters were assessed by using the intraclass correlation coefficient (ICC). The parameters of apparent diffusion coefficient (ADC), true diffusion coefficient (D), pseudo-diffusion coefficient (D*), and perfusion fraction (f) of nonhypervascular PNETs were compared with PDACs and hypervascular PNETs using the independent sample t test or the Mann-Whitney U test. The diagnostic performance was evaluated using receiver operating characteristic (ROC) analysis. RESULTS: All DWI parameters values showed good to excellent intra- and interobserver agreements (ICC = 0.743-0.873). Nonhypervascular PNETs had significantly lower ADC and D, but significantly higher f than PDACs (P = 0.005, P < 0.001 and P < 0.001, respectively). ADC, D and f of nonhypervascular PNETs were lower than hypervascular PNETs (P = 0.001, <0.001 and 0.093, respectively). D* of nonhypervascular PNETs showed no statistically significant differences with PDACs and hypervascular PNETs (P = 0.809 and 0.420). D showed a higher area under the curve (AUC), followed by ADC and f (AUC = 0.885, 0.665 and 0.740, respectively) in differentiating nonhypervascular PNETs from PDACs. CONCLUSION: Monoexponential and IVIM diffusion models are valuable to differentiate nonhypervascular PNETs from PDACs. D showed better performance than f and ADC.

Preoperative Multiparametric Quantitative Magnetic Resonance Imaging Correlates with Prognosis and Recurrence Patterns in Pancreatic Ductal Adenocarcinoma.

Magnetic resonance imaging (MRI) has been shown to be associated with prognosis in some tumors; however, the correlation in pancreatic ductal adenocarcinoma (PDAC) remains inconclusive. In this retrospective study, we ultimately included 136 patients and analyzed quantitative MRI parameters that are associated with prognosis and recurrence patterns in PDAC using survival analysis and competing risks models; all the patients have been operated on with histopathology and immunohistochemical staining for further evaluation. In intravoxel incoherent motion diffusion-weighted imaging (DWI), we found that pure-diffusion coefficient D value was an independent risk factor for overall survival (OS) (HR: 1.696, 95% CI: 1.003-2.869, p = 0.049) and recurrence-free survival (RFS) (HR: 2.066, 95% CI: 1.252-3.409, p = 0.005). A low D value (≤1.08 × 10-3 mm2/s) was significantly associated with a higher risk of local recurrence (SHR: 5.905, 95% CI: 2.107-16.458, p = 0.001). Subgroup analysis revealed that patients with high D and f values had significantly better outcomes with adjuvant chemotherapy. Distant recurrence patients in the high-D value group who received chemotherapy may significantly improve their OS and RFS. It was found that preoperative multiparametric quantitative MRI correlates with prognosis and recurrence patterns in PDAC. Diffusion coefficient D value can be used as a noninvasive biomarker for predicting prognosis and recurrence patterns in PDAC.

Giant cell tumour of the tendon sheath of the spine: clinical features and imaging findings.

OBJECTIVES: To review the clinical and imaging data of spinal giant cell tumour of the tendon sheath (GCTTS) to improve our understanding of the disease. METHODS: The imaging findings, clinicopathological features and clinical outcomes of 14 patients with pathologically confirmed spinal GCTTS were analysed retrospectively. RESULTS: All 14 patients had a single spinal lesion, including ten cervical vertebra lesions and four thoracic vertebra lesions. CT scan findings: The lesions showed osteolytic bone destruction and were centred on the facet joint, eroding the surrounding bone with a paravertebral soft tissue mass. MRI scan findings: all the lesions manifested predominantly as isointense or hypointense on T1-weighted imaging (T1WI). On T2-weighted imaging (T2WI), eight lesions were hypointense, and four were isointense. The remaining two lesions showed slight hyperintensity. The enhanced scans of eight lesions showed moderate to marked homogeneous or heterogeneous enhancement. PET/CT findings: Among the five patients who underwent PET/CT, three presented lesions with well-defined, sclerotic borders, and the uptake of 18F-FDG was markedly increased. One lesion showed an ill-defined border and an uneven increase in 18F-FDG uptake with an SUVmax value of 8.9. A recurrent lesion was only found on PET/CT 45 months after surgery and the SUVmax was 5.1. CONCLUSIONS: Spinal GCTTS is extremely rare. Osteolytic bone destruction in the area of the facet joint with a soft tissue mass and hypointensity on T2WI images are indicative of the spinal GCTTS. GCTTS shows high uptake of 18F-FDG, and PET/CT is helpful in detecting recurrent lesions.