Fostering confidence: evaluating inter-site reproducibility of Bingham-NODDI model measures using phantom and in-vivo acquisitions

The Neurite Orientation Dispersion and Density Imaging (NODDI) diffusion model offers insights into the intricated microstructure of the brain and a detailed view of the spatial organization of the neurites, revealing their alignment, the degree of their dispersion or organization, and the density of their distribution across distinct brain regions. The Bingham-NODDI model extends the Watson-NODDI formalism to address its limitations in capturing complex neurite structures such as bending and fanning, providing advanced insights into neurological conditions such as Multiple Sclerosis. However, the model still needs to undergo testing across different MRI scanner systems to ascertain its robustness. This study aims to evaluate its consistency by validating it using an MRI test object and in-vivo acquisitions, employing coefficient of variation analysis and inter-site comparability of the generated data. All acquisitions were performed on the 3T GE Signa Premier and Siemens MAGNETOM Prisma and on the 1.5T GE Artist. Results show consistency across scanners for the tensor model. Investigations into Mean Diffusivity values highlight a decrease with increasing echo times, requiring future further investigation. Complementing the phantom study, scans of healthy volunteers on GE Premier and Siemens Prisma affirm the Bingham-NODDI model's stability, with slight variations falling within acceptable margins of error for key metrics. In-depth analyses of the Orientation Dispersion Index, Tissue Volume Fraction, and Intra-neurite Volume Fraction underscore the model's reliability. The Bingham-NODDI model's demonstrated reliability serves as a solid foundation for detecting minor changes in brain microstructure over time, presenting a valuable tool for clinicians in both neuroscience and clinical fields. This study contributes to advancing multi-center studies by transcending individual equipment boundaries, paving the way for collaborative research and clinical applications.