Cross-sectional and longitudinal evaluation of white matter microstructure damage and cognitive correlations by automated fibre quantification in relapsing-remitting multiple sclerosis patients.

The purpose of this study was to characterize whole-brain white matter (WM) fibre tracts by automated fibre quantification (AFQ), capture subtle changes cross-sectionally and longitudinally in relapsing-remitting multiple sclerosis (RRMS) patients and explore correlations between these changes and cognitive performance A total of 114 RRMS patients and 71 healthy controls (HCs) were enrolled and follow-up investigations were conducted on 46 RRMS patients. Fractional anisotropy (FA), mean diffusion (MD), axial diffusivity (AD), and radial diffusivity (RD) at each node along the 20 WM fibre tracts identified by AFQ were investigated cross-sectionally and longitudinally in entire and pointwise manners. Partial correlation analyses were performed between the abnormal metrics and cognitive performance. At baseline, compared with HCs, patients with RRMS showed a widespread decrease in FA and increases in MD, AD, and RD among tracts. In the pointwise comparisons, more detailed abnormalities were localized to specific positions. At follow-up, although there was no significant difference in the entire WM fibre tract, there was a reduction in FA in the posterior portion of the right superior longitudinal fasciculus (R_SLF) and elevations in MD and AD in the anterior and posterior portions of the right arcuate fasciculus (R_AF) in the pointwise analysis. Furthermore, the altered metrics were widely correlated with cognitive performance in RRMS patients. RRMS patients exhibited widespread WM microstructure alterations at baseline and alterations in certain regions at follow-up, and the altered metrics were widely correlated with cognitive performance in RRMS patients, which will enhance our understanding of WM microstructure damage and its cognitive correlation in RRMS patients.

Radiomics derived from T2-FLAIR: the value of 2- and 3-classification tasks for different lesions in multiple sclerosis.

Background: White matter (WM) lesions can be classified into contrast enhancement lesions (CELs), iron rim lesions (IRLs), and non-iron rim lesions (NIRLs) based on different pathological mechanism in relapsing-remitting multiple sclerosis (RRMS). The application of radiomics established by T2-FLAIR to classify WM lesions in RRMS is limited, especially for 3-class classification among CELs, IRLs, and NIRLs. Methods: A total of 875 WM lesions (92 CELs, 367 IRLs, 416 NIRLs) were included in this study. The 2-class classification was only performed between IRLs and NIRLs. For the 2- and 3-class classification tasks, all the lesions were randomly divided into training and testing sets with a ratio of 8:2. We used least absolute shrinkage and selection operator (LASSO), reliefF algorithm, and mutual information (MI) for feature selection, then eXtreme gradient boosting (XGBoost), random forest (RF), and support vector machine (SVM) were used to establish discrimination models. Finally, the area under the curve (AUC), accuracy, sensitivity, specificity, and precision were used to evaluate the performance of the models. Results: For the 2-class classification model, LASSO classifier with RF model showed the best discrimination performance with the AUC of 0.893 (95% CI: 0.838-0.942), accuracy of 0.813, sensitivity of 0.833, specificity of 0.781, and precision of 0.851. However, the 3-class classification model of LASSO with XGBoost displayed the highest performance with the AUC of 0.920 (95% CI: 0.887-0.950), accuracy of 0.796, sensitivity of 0.839, specificity of 0.881, and precision of 0.846. Conclusions: Radiomics models based on T2-FLAIR images have the potential for discriminating among CELs, IRLs, and NIRLs in RRMS.

Multimodal Investigation of Deep Gray Matter Nucleus in Patients with Multiple Sclerosis and Their Clinical Correlations: A Multivariate Pattern Analysis Study.

Deep gray matter (DGM) nucleus are involved in patients with multiple sclerosis (MS) and are strongly associated with clinical symptoms. We used machine learning approach to further explore microstructural alterations in DGM of MS patients. One hundred and fifteen MS patients and seventy-one healthy controls (HC) underwent brain MRI. The fractional anisotropy (FA), mean diffusivity (MD), quantitative susceptibility value (QSV) and volumes of the caudate nucleus (CN), putamen (PT), globus pallidus (GP), and thalamus (TH) were measured. Multivariate pattern analysis, based on a machine-learning algorithm, was applied to investigate the most damaged regions. Partial correlation analysis was used to investigate the correlation between MRI quantitative metrics and clinical neurological scores. The area under the curve of FA-based classification model was 0.83, while they were 0.93 for MD and 0.81 for QSV. The Montreal cognitive assessment scores were correlated with the volume of the DGM and the expanded disability status scale scores were correlated with the MD of the GP and PT. The study results indicated that MS patients had involvement of DGM with the CN being the most affected. The atrophy of DGM in MS patients mainly affected cognitive function and the microstructural damage of DGM was mainly correlated with clinical disability.

Prediction of unenhanced lesion evolution in multiple sclerosis using radiomics-based models: a machine learning approach.

BACKGROUND: The volume change of multiple sclerosis (MS) lesion is related to its activity and can be used to assess disease progression. Therefore, the purpose of this study was to develop radiomics models for predicting the evolution of unenhanced MS lesions by using different kinds of machine learning algorithms and explore the optimal model. METHODS: In this prospective observation, 45 follow-up MR images obtained in 36 patients with MS (mean age 32.53±10.91; 23 women, 13 men) were evaluated. The lesions will be defined as interval activity and interval inactivity, respectively, based on the percentage of enlargement or reduction of the lesion >20% in the follow-up MR images. We extracted radiomic features of lesions on FLAIR images, and used recursive feature elimination (RFE), ReliefF algorithm and least absolute shrinkage and selection operator (LASSO) for feature selection, then three classification models including logistic regression, random forest and support vector machine (SVM) were used to build predictive models. The performance of the models were evaluated based on the sensitivity, specificity, precision, negative predictive value (NPV) and receiver operating characteristic curve (ROC) curves analyses. RESULTS: 135 interval inactivity lesions and 110 interval activity lesions were registered in our study. A total of 972 radiomics features were extracted, of which 265 were robust. The consistency and effectiveness of model performance were compared and verified by different combinations of feature selection and machine learning methods in different K-fold cross-validation strategies where K ranges from 5 to 10, thus demonstrating the stability and robustness. SVM classifier with ReliefF algorithm had the best prediction performance with an average accuracy of 0.827, sensitivity of 0.809, specificity of 0.841, precision of 0.921, NPV of 0.948 and the areas under the ROC curves (AUC) of 0.857 (95% CI: 0.812-0.902) in the cohorts. CONCLUSION: The results demonstrated that the radiomics-based machine learning model has potential in predicting the evolution of MS lesions.