Standard Model Imaging for Characterizing Multiple Sclerosis Lesion Types: A Lesion-Focused Analysis Compared with Diffusion Tensor Imaging

PurposeTo characterize microstructural alterations across distinct white matter tissue classes in MS using Standard Model Imaging (SMI), and to place its performance in context relative to conventional diffusion tensor imaging (DTI). MethodsDTI and SMI were applied to treatment-naive individuals at early stages of MS, including patients with MS and healthy controls. Over 3,602 manually delineated regions of interest were classified into normal white matter (NWM), normal-appearing white matter (NAWM), T2-hyperintense lesions, and chronic black holes (cBHs) differences were assessed using linear mixed-effects models with false discovery rate correction. Discriminative performance was evaluated using receiver operating characteristic (ROC) analysis within a generalized linear mixed modeling framework for individual parameters and multivariate DTI, SMI, and combined DTI+SMI models. ResultsBoth DTI and SMI metrics demonstrated widespread and significant differences across tissue classes. Robust discriminative performance was observed for lesion-NWM and lesion-NAWM comparisons (AUC > 0.8), whereas discrimination between NAWM and NWM and between cBHs and T2-lesions was limited (AUC [≤] 0.66). In terms of model performance, SMI achieved slightly higher AUC values than DTI across most contrasts, while the combined DTI+SMI model consistently provided the highest diagnostic performance. ROI-based analyses revealed additional SMI alterations, including changes in extra-axonal parallel diffusivity, not consistently reported in prior studies. ConclusionDTI and SMI metrics are sensitive to microstructural abnormalities across a broad spectrum of white matter tissue classes in MS, capturing both lesion-related damage and more subtle alterations extending into NAWM. While discriminative performance varies by tissue contrast, integrating DTI and SMI provides complementary information and modestly improves diagnostic performance, supporting a multi-model diffusion MRI approach for comprehensive characterization of MS-related white matter pathology