Comparison of Neurite Orientation Dispersion and Density Imaging and Two-Compartment Spherical Mean Technique Parameter Maps in Multiple Sclerosis.

Daniel Johnson,Antonio Ricciardi,Baris Kanber,Wallace Brownlee,Claudia A. M. Gandini Wheeler-Kingshott,Ferran Prados,Francesco Grussu,Sara Collorone,Enrico Kaden,Ahmed Toosy,Daniel C. Alexander,Olga Ciccarelli

doi:10.3389/fneur.2021.662855

Abstract

Background: Neurite orientation dispersion and density imaging (NODDI) and the spherical mean technique (SMT) are diffusion MRI methods providing metrics with sensitivity to similar characteristics of white matter microstructure. There has been limited comparison of changes in NODDI and SMT parameters due to multiple sclerosis (MS) pathology in clinical settings.Purpose: To compare group-wise differences between healthy controls and MS patients in NODDI and SMT metrics, investigating associations with disability and correlations with diffusion tensor imaging (DTI) metrics.Methods: Sixty three relapsing-remitting MS patients were compared to 28 healthy controls. NODDI and SMT metrics corresponding to intracellular volume fraction (vin), orientation dispersion (ODI and ODE), diffusivity (D) (SMT only) and isotropic volume fraction (viso) (NODDI only) were calculated from diffusion MRI data, alongside DTI metrics (fractional anisotropy, FA; axial/mean/radial diffusivity, AD/MD/RD). Correlations between all pairs of MRI metrics were calculated in normal-appearing white matter (NAWM). Associations with expanded disability status scale (EDSS), controlling for age and gender, were evaluated. Patient-control differences were assessed voxel-by-voxel in MNI space controlling for age and gender at the 5% significance level, correcting for multiple comparisons. Spatial overlap of areas showing significant differences were compared using Dice coefficients.Results: NODDI and SMT show significant associations with EDSS (standardised beta coefficient −0.34 in NAWM and −0.37 in lesions for NODDI vin; 0.38 and −0.31 for SMT ODE and vin in lesions; p < 0.05). Significant correlations in NAWM are observed between DTI and NODDI/SMT metrics. NODDI vin and SMT vin strongly correlated (r = 0.72, p < 0.05), likewise NODDI ODI and SMT ODE (r = −0.80, p < 0.05). All DTI, NODDI and SMT metrics detect widespread differences between patients and controls in NAWM (12.57% and 11.90% of MNI brain mask for SMT and NODDI vin, Dice overlap of 0.42).Data Conclusion: SMT and NODDI detect significant differences in white matter microstructure between MS patients and controls, concurring on the direction of these changes, providing consistent descriptors of tissue microstructure that correlate with disability and show alterations beyond focal damage. Our study suggests that NODDI and SMT may play a role in monitoring MS in clinical trials and practice.

Highlights

Over the last 20 years, diffusion tensor imaging (DTI) [1] has established itself as a modality of choice for imaging the cerebral white matter (WM) with a view to gaining an understanding of its microstructure in health and disease
In this study we have considered DTI indices, namely fractional anisotropy (FA), AD, RD and mean diffusivity (MD), as these provide a well-established reference to which the performance of Neurite orientation dispersion and density imaging (NODDI) and spherical mean technique (SMT) can be compared
For intraneurite volume fraction vin, our results point toward reduction in this metric in both lesions and normal-appearing white matter (NAWM) in RRMS patients compared with healthy controls

Summary

Introduction

Over the last 20 years, diffusion tensor imaging (DTI) [1] has established itself as a modality of choice for imaging the cerebral white matter (WM) with a view to gaining an understanding of its microstructure in health and disease. In white matter, diffusivity is generally significantly larger parallel to axons rather than perpendicular to them producing a large FA This anisotropy can be reduced either by a reduction in the density of axons, an increase in their orientation dispersion, or a combination of both. Neurite orientation dispersion and density imaging (NODDI) and the spherical mean technique (SMT) are diffusion MRI methods providing metrics with sensitivity to similar characteristics of white matter microstructure. Purpose: To compare group-wise differences between healthy controls and MS patients in NODDI and SMT metrics, investigating associations with disability and correlations with diffusion tensor imaging (DTI) metrics

Methods

Results

Discussion

Conclusion