Title : Radiomics evaluation for the early detection of multiple sclerosis using T2-weighted MRI
Abstract:
Introduction: Multiple sclerosis (MS) is a chronic inflammatory disease of the central nervous system that benefits significantly from early diagnosis and treatment. This study evaluates the efficacy of radiomics features extracted from T2-weighted magnetic resonance imaging (MRI) for the early detection of MS.
Methods: We retrospectively analyzed T2-weighted MRI scans from 280 participants (163 with early MS, and 117 age-matched healthy controls) collected across neurological center. A comprehensive radiomics pipeline extracted 1,358 quantitative features from regions of interest including T2 hyperintense lesions, periventricular white matter, subcortical white matter, and deep gray matter structures. Feature extraction encompassed first-order statistics, shape-based metrics, textural properties, and wavelet transformations. Feature selection using maximum relevance minimum redundancy (mRMR) identified 53 radiomics features with optimal discriminatory power.
Results: A support vector machine classifier trained on these features achieved an area under the receiver operating characteristic curve (AUC) of 0.94 (95% CI: 0.91-0.97) for distinguishing early MS from healthy controls and 0.87 (95% CI: 0.83-0.91) for discriminating RIS from healthy controls in the independent validation cohort. Texture features capturing lesion heterogeneity and periventricular white matter characteristics demonstrated the highest discriminative value. The radiomics signature significantly outperformed conventional lesion count and volume metrics (AUC 0.94 vs. 0.82, p<0.001), particularly in cases with low lesion burden. Longitudinal analysis in a subset of 96 participants revealed that specific radiomics features showed significant alterations 6-14 months before patients met the clinical criteria for MS diagnosis. Integration of radiomics features with clinical and demographic data further improved the model performance (AUC 0.96, 95% CI: 0.94-0.98).
Conclusion: These findings suggest that radiomics analysis of T2-weighted MRI provides sensitive imaging biomarkers for early detection of MS, potentially enabling earlier therapeutic intervention and more precise risk stratification for patients with initial demyelinating events.
Keywords: Multiple Sclerosis, Radiomics, Early Detection, Machine Learning, Neuroimaging Biomarkers
