Synthesis of gadolinium-enhanced glioma images on multisequence magnetic resonance images using contrastive learning.

BACKGROUND: Gadolinium-based contrast agents are commonly used in brain magnetic resonance imaging (MRI), however, they cannot be used by patients with allergic reactions or poor renal function. For long-term follow-up patients, gadolinium deposition in the body can cause nephrogenic systemic fibrosis and other potential risks. PURPOSE: Developing a new method of enhanced image synthesis based on the advantages of multisequence MRI has important clinical value for these patients. In this paper, an end-to-end synthesis model structure similarity index measure (SSIM)-based Dual Constrastive Learning with Attention (SDACL) based on contrastive learning is proposed to synthesize contrast-enhanced T1 (T1ce) using three unenhanced MRI images of T1, T2, and Flair in patients with glioma. METHODS: The model uses the attention-dilation generator to enlarge the receptive field by expanding the residual blocks and to strengthen the feature representation and context learning of multisequence MRI. To enhance the detail and texture performance of the imaged tumor area, a comprehensive loss function combining patch-level contrast loss and structural similarity loss is created, which can effectively suppress noise and ensure the consistency of synthesized images and real images. RESULTS: The normalized root-mean-square error (NRMSE), peak signal-to-noise ratio (PSNR), and SSIM of the model on the independent test set are 0.307  ± $\pm$  0.12, 23.337  ± $\pm$  3.21, and 0.881  ± $\pm$  0.05, respectively. CONCLUSIONS: Results show this method can be used for the multisequence synthesis of T1ce images, which can provide valuable information for clinical diagnosis.

Universal multi-factor feature selection method for radiomics-based brain tumor classification.

Brain tumor mortality is high, and accurate classification before treatment can improve patient prognosis. Radiomics, which extracts numerous features from medical images, has been widely applied in brain tumor classification studies. Feature selection (FS) is a critical step in radiomics because it reduces redundant information and enhances classification performance. However, the lack of universal FS methods limits the development of radiomics-based brain tumor classification studies. To address this issue, we summarize the characteristics of the FS methods used in related studies and propose a universal method based on three selection factors called triple-factor cascaded selection (TFCS). Particularly, these factors correspond to the correlation between features and task labels, interdependence among features, and role of features in the model. The TFCS method divides FS into two steps. First, it utilizes mutual information to select features that are strongly correlated with the task and contain less redundant information. Recursive feature elimination is then employed to obtain the subset with the best classification performance. To validate the universality of the TFCS, we conducted experiments on seven datasets containing 13 brain tumor classification tasks and evaluated the overall performance using five types of indicators. Results: TFCS exhibited excellent overall performance for all tasks. Compared to the 13 related methods, it takes less time, has moderate parsimony, the best classification performance, adaptability, and stability, and shows better universality. Our study demonstrates that the reasonable utilization of multiple factors can enhance FS performance and provide new insights for future method design.

Multi-user multi-objective computation offloading for medical image diagnosis.

Computation offloading has effectively solved the problem of terminal devices computing resources limitation in hospitals by shifting the medical image diagnosis task to the edge servers for execution. Appropriate offloading strategies for diagnostic tasks are essential. However, the risk awareness of each user and the multiple expenses associated with processing tasks have been ignored in prior works. In this article, a multi-user multi-objective computation offloading for medical image diagnosis is proposed. First, the prospect theoretic utility function of each user is designed considering the delay, energy consumption, payment, and risk awareness. Second, the computation offloading problem including the above factors is defined as a distributed optimization problem, which with the goal of maximizing the utility of each user. The distributed optimization problem is then transformed into a non-cooperative game among the users. The exact potential game proves that the non-cooperative game has Nash equilibrium points. A low-complexity computation offloading algorithm based on best response dynamics finally is proposed. Detailed numerical experiments demonstrate the impact of different parameters and convergence in the algorithm on the utility function. The result shows that, compare with four benchmarks and four heuristic algorithms, the proposed algorithm in this article ensures a faster convergence speed and achieves only a 1.14% decrease in the utility value as the number of users increases.

Deep residual-SVD network for brain image registration.

Objective.Medical image registration aims to find the deformation field that can align two images in a spatial position. A medical image registration method based on U-Net architecture has been proposed currently. However, U-Net architecture has few training parameters, which leads to weak learning ability, and it ignores the adverse effects of image noise on the registration accuracy. The article aims at addressing the problem of weak network learning ability and the adverse effects of noisy images on registration.Approach.Here we propose a novel unsupervised 3D brain image registration framework, which introduces the residual unit and singular value decomposition (SVD) denoising layer on the U-Net architecture. Residual unit solves the problem of network degradation, that is, registration accuracy becomes saturated and then degrades rapidly with the increase in network depth. SVD denoising layer uses the estimated model order for SVD-based low-rank image reconstruction. we use Akaike information criterion to estimate the appropriate model order, which is used to remove noise components. We use the exponential linear unit (ELU) as the activation function, which is more robust to noise than other peers.Main results.The proposed method is evaluated on the publicly available brain MRI datasets: Mindboggle101 and LPBA40. Experimental results demonstrate our method outperforms several state-of-the-art methods for the metric of Dice Score. The mean number of folding voxels and registration time are comparable to state-of-the-art methods.Significance.This study shows that Deep Residual-SVD Network can improve registration accuracy. This study also demonstrate that the residual unit can enhance the learning ability of the network, the SVD denoising layer can denoise effectively, and the ELU is more robust to noise.

Multiparametric Magnetic Resonance Imaging Information Fusion Using Graph Convolutional Network for Glioma Grading.

Accurate preoperative glioma grading is essential for clinical decision-making and prognostic evaluation. Multiparametric magnetic resonance imaging (mpMRI) serves as an important diagnostic tool for glioma patients due to its superior performance in describing noninvasively the contextual information in tumor tissues. Previous studies achieved promising glioma grading results with mpMRI data utilizing a convolutional neural network (CNN)-based method. However, these studies have not fully exploited and effectively fused the rich tumor contextual information provided in the magnetic resonance (MR) images acquired with different imaging parameters. In this paper, a novel graph convolutional network (GCN)-based mpMRI information fusion module (named MMIF-GCN) is proposed to comprehensively fuse the tumor grading relevant information in mpMRI. Specifically, a graph is constructed according to the characteristics of mpMRI data. The vertices are defined as the glioma grading features of different slices extracted by the CNN, and the edges reflect the distances between the slices in a 3D volume. The proposed method updates the information in each vertex considering the interaction between adjacent vertices. The final glioma grading is conducted by combining the fused information in all vertices. The proposed MMIF-GCN module can introduce an additional nonlinear representation learning step in the process of mpMRI information fusion while maintaining the positional relationship between adjacent slices. Experiments were conducted on two datasets, that is, a public dataset (named BraTS2020) and a private one (named GliomaHPPH2018). The results indicate that the proposed method can effectively fuse the grading information provided in mpMRI data for better glioma grading performance.

Short-Axis PET Image Quality Improvement by Attention CycleGAN Using Total-Body PET.

The quality of positron emission tomography (PET) imaging is positively correlated with scanner sensitivity, which is closely related to the axial field of view (FOV). Conventional short-axis PET scanners (200-350 mm FOV) reduce the imaging quality during fast scanning (2-3 minutes) due to the limitation of FOV, which reduce the reliability of diagnosis. To overcome hardware limitations and improve the image quality of short-axis PET scanners, we propose a supervised deep learning model, CycleAGAN, which is based on a cycle-consistent adversarial network (CycleGAN). We introduced the attention mechanism into the generator and focus on channel and spatial representative features and supervised learning using pairs of data to maintain the spatial consistency of the generated images with the ground truth. The imaging information of 386 patients from Henan Provincial People's Hospital was prospectively included as the dataset in this study. The training data come from the total-body PET scanner uEXPLORER. The proposed CycleAGAN is compared with traditional gray-level-based methods and learning-based methods. The results confirm that CycleAGAN achieved the best results on SSIM and NRMSE and achieved the closest distribution to ground truth in expert rating. The proposed method is not only able to improve the image quality of PET scanners with 320 mm FOV but also achieved good results on shorter FOV scanners. Patients and radiologists can benefit from the computer-aided diagnosis (CAD) system integrated with CycleAGAN.

Bayesian Fully Convolutional Networks for Brain Image Registration.

The purpose of medical image registration is to find geometric transformations that align two medical images so that the corresponding voxels on two images are spatially consistent. Nonrigid medical image registration is a key step in medical image processing, such as image comparison, data fusion, target recognition, and pathological change analysis. Existing registration methods only consider registration accuracy but largely neglect the uncertainty of registration results. In this work, a method based on the Bayesian fully convolutional neural network is proposed for nonrigid medical image registration. The proposed method can generate a geometric uncertainty map to calculate the uncertainty of registration results. This uncertainty can be interpreted as a confidence interval, which is essential for judging whether the source data are abnormal. Moreover, the proposed method introduces group normalization, which is conducive to the network convergence of the Bayesian neural network. Some representative learning-based image registration methods are compared with the proposed method on different image datasets. Experimental results show that the registration accuracy of the proposed method is better than that of the methods, and its antifolding performance is comparable to that of fast image registration and VoxelMorph. Furthermore, the proposed method can evaluate the uncertainty of registration results.

Identification of Arrhythmia by Using a Decision Tree and Gated Network Fusion Model.

In recent years, deep learning (DNN) based methods have made leapfrogging level breakthroughs in detecting cardiac arrhythmias as the cost effectiveness of arithmetic power, and data size has broken through the tipping point. However, the inability of these methods to provide a basis for modeling decisions limits clinicians' confidence on such methods. In this paper, a Gate Recurrent Unit (GRU) and decision tree fusion model, referred to as (T-GRU), was designed to explore the problem of arrhythmia recognition and to improve the credibility of deep learning methods. The fusion model multipathway processing time-frequency domain featured the introduction of decision tree probability analysis of frequency domain features, the regularization of GRU model parameters and weight control to improve the decision tree model output weights. The MIT-BIH arrhythmia database was used for validation. Results showed that the low-frequency band features dominated the model prediction. The fusion model had an accuracy of 98.31%, sensitivity of 96.85%, specificity of 98.81%, and precision of 96.73%, indicating its high reliability and clinical significance.

Fproi-GAN with Fused Regional Features for the Synthesis of High-Quality Paired Medical Images.

The use of medical image synthesis with generative adversarial networks (GAN) is effective for expanding medical samples. The structural consistency between the synthesized and actual image is a key indicator of the quality of the synthesized image, and the region of interest (ROI) of the synthesized image is related to its usability, and these parameters are the two key issues in image synthesis. In this paper, the fusion-ROI patch GAN (Fproi-GAN) model was constructed by incorporating a priori regional feature based on the two-stage cycle consistency mechanism of cycleGAN. This model has improved the tissue contrast of ROI and achieved the pairwise synthesis of high-quality medical images and their corresponding ROIs. The quantitative evaluation results in two publicly available datasets, INbreast and BRATS 2017, show that the synthesized ROI images have a DICE coefficient of 0.981 ± 0.11 and a Hausdorff distance of 4.21 ± 2.84 relative to the original images. The classification experimental results show that the synthesized images can effectively assist in the training of machine learning models, improve the generalization performance of prediction models, and improve the classification accuracy by 4% and sensitivity by 5.3% compared with the cycleGAN method. Hence, the paired medical images synthesized using Fproi-GAN have high quality and structural consistency with real medical images.

SAP-cGAN: Adversarial learning for breast mass segmentation in digital mammogram based on superpixel average pooling.

PURPOSE: Breast mass segmentation is a prerequisite step in the use of computer-aided tools designed for breast cancer diagnosis and treatment planning. However, mass segmentation remains challenging due to the low contrast, irregular shapes, and fuzzy boundaries of masses. In this work, we propose a mammography mass segmentation model for improving segmentation performance. METHODS: We propose a mammography mass segmentation model called SAP-cGAN, which is based on an improved conditional generative adversarial network (cGAN). We introduce a superpixel average pooling layer into the cGAN decoder, which utilizes superpixels as a pooling layout to improve boundary segmentation. In addition, we adopt a multiscale input strategy to enable the network to learn scale-invariant features with increased robustness. The performance of the model is evaluated with two public datasets: CBIS-DDSM and INbreast. Moreover, ablation analysis is conducted to evaluate further the individual contribution of each block to the performance of the network. RESULTS: Dice and Jaccard scores of 93.37% and 87.57%, respectively, are obtained for the CBIS-DDSM dataset. The Dice and Jaccard scores for the INbreast dataset are 91.54% and 84.40%, respectively. These results indicate that our proposed model outperforms current state-of-the-art breast mass segmentation methods. The superpixel average pooling layer and multiscale input strategy has improved the Dice and Jaccard scores of the original cGAN by 7.8% and 12.79%, respectively. CONCLUSIONS: Adversarial learning with the addition of a superpixel average pooling layer and multiscale input strategy can encourage the Generator network to generate masks with increased realism and improve breast mass segmentation performance through the minimax game between the Generator network and Discriminator network.

A New Multichannel Parallel Network Framework for the Special Structure of Multilead ECG.

Electrocardiogram (ECG) contains the rhythmic features of continuous heartbeat and morphological features of ECG waveforms and varies among different diseases. Based on ECG signal features, we propose a combination of multiple neural networks, the multichannel parallel neural network (MLCNN-BiLSTM), to explore feature information contained in ECG. The MLCNN channel is used in extracting the morphological features of ECG waveforms. Compared with traditional convolutional neural network (CNN), the MLCNN can accurately extract strong relevant information on multilead ECG while ignoring irrelevant information. It is suitable for the special structures of multilead ECG. The Bidirectional Long Short-Term Memory (BiLSTM) channel is used in extracting the rhythmic features of ECG continuous heartbeat. Finally, by initializing the core threshold parameters and using the backpropagation algorithm to update automatically, the weighted fusion of the temporal-spatial features extracted from multiple channels in parallel is used in exploring the sensitivity of different cardiovascular diseases to morphological and rhythmic features. Experimental results show that the accuracy rate of multiple cardiovascular diseases is 87.81%, sensitivity is 86.00%, and specificity is 87.76%. We proposed the MLCNN-BiLSTM neural network that can be used as the first-round screening tool for clinical diagnosis of ECG.

Mammography-based radiomics nomogram: a potential biomarker to predict axillary lymph node metastasis in breast cancer.

OBJECTIVE: To establish a radiomics nomogram by integrating clinical risk factors and radiomics features extracted from digital mammography (MG) images for pre-operative prediction of axillary lymph node (ALN) metastasis in breast cancer. METHODS: 216 patients with breast cancer lesions confirmed by surgical excision pathology were divided into the primary cohort (n = 144) and validation cohort (n = 72). Radiomics features were extracted from craniocaudal (CC) view of mammograms, and radiomics features selection were performed using the methods of ANOVA F-value and least absolute shrinkage and selection operator; then a radiomics signature was constructed with the method of support vector machine. Multivariate logistic regression analysis was used to establish a radiomics nomogram based on the combination of radiomics signature and clinical factors. The C-index and calibration curves were derived based on the regression analysis both in the primary and validation cohorts. RESULTS: 95 of 216 patients were confirmed with ALN metastasis by pathology, and 52 cases were diagnosed as ALN metastasis based on MG-reported criteria. The sensitivity, specificity, accuracy and AUC (area under the receiver operating characteristic curve of MG-reported criteria were 42.7%, 90.8%, 24.1% and 0.666 (95% confidence interval: 0.591-0.741]. The radiomics nomogram, comprising progesterone receptor status, molecular subtype and radiomics signature, showed good calibration and better favorite performance for the metastatic ALN detection (AUC 0.883 and 0.863 in the primary and validation cohorts) than each independent clinical features (AUC 0.707 and 0.657 in the primary and validation cohorts) and radiomics signature (AUC 0.876 and 0.862 in the primary and validation cohorts). CONCLUSION: The MG-based radiomics nomogram could be used as a non-invasive and reliable tool in predicting ALN metastasis and may facilitate to assist clinicians for pre-operative decision-making. ADVANCES IN KNOWLEDGE: ALN status remains among the most important breast cancer prognostic factors and is essential for making treatment decisions. However, the value of detecting metastatic ALN by MG is very limited. The studies on pre-operative ALN metastasis prediction using the method of MG-based radiomics in breast cancer are very few. Therefore, we studied whether MG-based radiomics nomogram could be used as a predictive biomarker for the detection of metastatic ALN.

A Novel Algorithm for Breast Mass Classification in Digital Mammography Based on Feature Fusion.

Prompt diagnosis of benign and malignant breast masses is essential for early breast cancer screening. Convolutional neural networks (CNNs) can be used to assist in the classification of benign and malignant breast masses. A persistent problem in current mammography mass classification via CNN is the lack of local-invariant features, which cannot effectively respond to geometric image transformations or changes caused by imaging angles. In this study, a novel model that trains both texton representation and deep CNN representation for mass classification tasks is proposed. Rotation-invariant features provided by the maximum response filter bank are incorporated with the CNN-based classification. The fusion after implementing the reduction approach is used to address the deficiencies of CNN in extracting mass features. This model is tested on public datasets, CBIS-DDSM, and a combined dataset, namely, mini-MIAS and INbreast. The fusion after implementing the reduction approach on the CBIS-DDSM dataset outperforms that of the other models in terms of area under the receiver operating curve (0.97), accuracy (94.30%), and specificity (97.19%). Therefore, our proposed method can be integrated with computer-aided diagnosis systems to achieve precise screening of breast masses.

Preoperative Prediction of Axillary Lymph Node Metastasis in Breast Carcinoma Using Radiomics Features Based on the Fat-Suppressed T2 Sequence.

RATIONALE AND OBJECTIVES: To investigate the value of radiomics method based on the fat-suppressed T2 sequence for preoperative predicting axillary lymph node (ALN) metastasis in breast carcinoma. MATERIALS AND METHODS: The data of 329 invasive breast cancer patients were divided into the primary cohort (n = 269) and validation cohort (n = 60). Radiomics features were extracted from the fat-suppressed T2-weighted images on breast MRI, and ALN metastasis-related radiomics feature selection was performed using Mann-Whitney U-test and support vector machines with recursive feature elimination; then a radiomics signature was constructed by linear support vector machine. The predictive models were constructed using a linear regression model based on the clinicopathologic factors and radiomics signature, and nomogram was used for a visual prediction of the combined model. The predictive performances are evaluated with the sensitivity, specificity, accuracy, and area under the receiver operating characteristic curve. RESULTS: A total of 647 radiomics features were extracted from each patient. About 23 ALN metastasis-related radiomics features were selected to construct the radiomics signature, including 17 texture features, 5 first-order statistical features, and one shape feature; patient age, tumor size, HER2 status, and vascular cancer thrombus accompanied or not were selected to construct the cilinicopathologic feature model. The sensitivity, specificity, accuracy, and are under the curve value of radiomics signature, clinicopathologic feature model, and the nomogram were 65.22%, 81.08%, 75.00%, and 0.819 (95% confidence interval [CI]: 0.776-0.861), 30.44%, 81.08%, 61.67%, and 0.605 (95% CI: 0.571-0.624) and 60.87%, 89.19%, 78.33%, and 0.810 (95% CI: 0.761-0.855), respectively. CONCLUSION: Radiomics methods based on the fat-suppressed T2 sequence and the nomogram are helpful for preoperative accurate predicting ALN metastasis.

Identification of Cognitive Dysfunction in Patients with T2DM Using Whole Brain Functional Connectivity.

Majority of type 2 diabetes mellitus (T2DM) patients are highly susceptible to several forms of cognitive impairments, particularly dementia. However, the underlying neural mechanism of these cognitive impairments remains unclear. We aimed to investigate the correlation between whole brain resting state functional connections (RSFCs) and the cognitive status in 95 patients with T2DM. We constructed an elastic net model to estimate the Montreal Cognitive Assessment (MoCA) scores, which served as an index of the cognitive status of the patients, and to select the RSFCs for further prediction. Subsequently, we utilized a machine learning technique to evaluate the discriminative ability of the connectivity pattern associated with the selected RSFCs. The estimated and chronological MoCA scores were significantly correlated with R = 0.81 and the mean absolute error (MAE) = 1.20. Additionally, cognitive impairments of patients with T2DM can be identified using the RSFC pattern with classification accuracy of 90.54% and the area under the receiver operating characteristic (ROC) curve (AUC) of 0.9737. This connectivity pattern not only included the connections between regions within the default mode network (DMN), but also the functional connectivity between the task-positive networks and the DMN, as well as those within the task-positive networks. The results suggest that an RSFC pattern could be regarded as a potential biomarker to identify the cognitive status of patients with T2DM.

A Non-invasive Radiomic Method Using 18F-FDG PET Predicts Isocitrate Dehydrogenase Genotype and Prognosis in Patients With Glioma.

Purpose: We aimed to analyze 18F-fluorodeoxyglucose positron emission tomography (18F-FDG PET) images via the radiomic method to develop a model and validate the potential value of features reflecting glioma metabolism for predicting isocitrate dehydrogenase (IDH) genotype and prognosis. Methods: PET images of 127 patients were retrospectively analyzed. A series of quantitative features reflecting the metabolic heterogeneity of the tumors were extracted, and a radiomic signature was generated using the support vector machine method. A combined model that included clinical characteristics and the radiomic signature was then constructed by multivariate logistic regression to predict the IDH genotype status, and the model was evaluated and verified by receiver operating characteristic (ROC) curves and calibration curves. Finally, Kaplan-Meier curves and log-rank tests were used to analyze overall survival (OS) according to the predicted result. Results: The generated radiomic signature was significantly associated with IDH genotype (p < 0.05) and could achieve large areas under the ROC curve of 0.911 and 0.900 on the training and validation cohorts, respectively, with the incorporation of age and type of tumor metabolism. The good agreement of the calibration curves in the validation cohort further validated the efficacy of the constructed model. Moreover, the predicted results showed a significant difference in OS between high- and low-risk groups (p < 0.001). Conclusions: Our results indicate that the 18F-FDG metabolism-related features could effectively predict the IDH genotype of gliomas and stratify the OS of patients with different prognoses.

Radiomics analysis of placenta on T2WI facilitates prediction of postpartum haemorrhage: A multicentre study.

BACKGROUND: Identification of pregnancies with postpartum haemorrhage (PPH) antenatally rather than intrapartum would aid delivery planning, facilitate transfusion requirements and decrease maternal complications. MRI has been increasingly used for placenta evaluation. Here, we aim to build a nomogram incorporating both clinical and radiomic features of placenta to predict the risk for PPH in pregnancies during caesarian delivery (CD). METHODS: A total of 298 pregnant women were retrospectively enrolled from Henan Provincial People's Hospital (training cohort: n = 207) and from The Third Affiliated Hospital of Zhengzhou University (external validation cohort: n = 91). These women were suspected with placenta accreta spectrum (PAS) disorders and underwent MRI for placenta evaluation. All of them underwent CD and were singleton. PPH was defined as more than 1000 mL estimated blood loss (EBL) during CD. Radiomic features were selected based on their correlations with EBL. Radiomic, clinical, radiological, clinicoradiological and clinicoradiomic models were built to predict the risk of PPH for each patient. The model with the best prediction performance was validated with its discrimination ability, calibration curve and clinical application. FINDINGS: Thirty-five radiomic features showed strong correlation with EBL. The clinicoradiomic model resulted in the best discrimination ability for risk prediction of PPH, with AUC of 0.888 (95% CI, 0.844-0.933) and 0.832 (95% CI, 0.746-0.913), sensitivity of 91.2% (95% CI, 85.8%-96.7%) and 97.6% (95% CI, 92.7%-100%) in the training and validation cohort respectively. For patients with severe PPH (EBL more than 2000 mL), 53 out of 55 pregnancies (96.4%) in the training cohort and 18 out of 18 (100%) pregnancies in the validation cohort were identified by the clinicoradiomic model. The model performed better in patients without placenta previa (PP) than in patients with PP, with AUC of 0.983 compared with 0.867, sensitivity of 100% compared with 90.8% in the training cohort, AUC of 0.832 compared with 0.815, sensitivity of 97.6% compared with 97.2% in the validation cohort. INTERPRETATION: The clinicoradiomic model incorporating both prenatal clinical factors and radiomic signature of placenta on T2WI showed good performance for risk prediction of PPH. The predictive model can identify severe PPH with high sensitivity and can be applied in patients with and without PP.

Development and validation of an MRI-based radiomic signature for the preoperative prediction of treatment response in patients with invasive functional pituitary adenoma.

PURPOSE: The preoperative prediction of treatment response is important for determining individual treatment strategies for invasive functional pituitary adenoma (IFPA). This study aimed to develop and validate a magnetic resonance imaging (MRI)-based radiomic signature for preoperative prediction of treatment response in IFPA. METHOD: One hundred and sixty-three patients with IFPA were enrolled and divided into primary (n = 108) and validation cohorts (n = 55) according to time point. IFPA patients were divided into remission and non-remission according to postoperative hormone levels. Radiomic features were extracted from their MR images and a radiomic signature was built using a support vector machine. Subsequently, multivariable logistic regression analysis was used to select the most informative clinical features, and a radiomic model incorporating the radiomic signature and selected clinical features was constructed and used as the final predictive model. RESULTS: Seven radiomic features were selected to construct the radiomic signature, which achieved an area under the curve (AUC) of 0.834 and 0.808 on the primary and validation cohorts respectively. The radiomic model incorporating the radiomic signature and Knosp grade showed good discrimination abilities and calibration, with AUCs of 0.832 and 0.811 for the primary and validation cohorts respectively. The radiomic signature and radiomic model better estimated the treatment responses of patients with IFPA than our clinical features model. Decision curve analysis showed the radiomic model was clinically useful. CONCLUSIONS: This radiomic model may help neurosurgeons predict the treatment responses of patients with IFPA before surgery and determine individual treatment strategies.

Radiomic analysis of multiparametric magnetic resonance imaging for differentiating skull base chordoma and chondrosarcoma.

PURPOSE: Patients with skull base chordoma and chondrosarcoma have different prognoses and are not readily differentiated preoperatively on imaging. Multiparametric magnetic resonance imaging (MRI) is a routine diagnostic tool that can noninvasively characterize the salient characteristics of tumors. In the present study, we developed and validated a preoperative multiparametric MRI-based radiomic signature for differentiating these tumors. METHOD: This retrospective study enrolled 210 patients and consecutively divided them into the primary and validation cohorts. A total of 1941 radiomic features were acquired from preoperative T1-weighted imaging, T2-weighted imaging and contrast-enhanced T1-weighted imaging for each patient. The most discriminative features were selected by minimum-redundancy maximum-relevancy and recursive feature elimination algorithms in the primary cohort. The multiparametric and single-sequence MRI signatures were constructed with the selected features using a support vector machine model in the primary cohort. The ability of the novel radiomic signatures to differentiate chordoma from chondrosarcoma were assessed using receiver operating characteristic curve analysis in the validation cohort. RESULTS: The multiparametric radiomic signature, which consisted of 11 selected features, reached an area under the receiver operating characteristic curve of 0.9745 and 0.8720 in the primary and validation cohorts, respectively. Moreover, compared with each single-sequence MRI signature, the multiparametric radiomic signature exhibited better classification performance with significant improvement (p <  0.05, Delong's test) in the primary cohorts. CONCLUSION: By combining features from three MRI sequences, the multiparametric radiomics signature can accurately and robustly differentiate skull base chordoma from chondrosarcoma.