RobNODDI: Robust NODDI Parameter Estimation with Adaptive Sampling under Continuous Representation

TL;DR

RobNODDI tackles the critical issue of poor generalization in diffusion MRI NODDI parameter estimation when testing diffusion directions differ from training. It introduces adaptive sampling and a SH-based continuous representation to convert DWI into a stable, direction-agnostic feature set, enabling a flexible network (built on MESC-SD) to map SH coefficients to NODDI parameters. Empirical results on the HCP dataset show improved generalization and robustness in $V_{icvf}$, $V_{iso}$, and $OD$ estimation, outperforming existing DL methods under RS and benefiting from SH input. The approach offers enhanced clinical applicability by reducing sensitivity to acquisition schemes and enabling more reliable microstructural assessments in diverse diffusion MRI protocols.

Abstract

Neurite Orientation Dispersion and Density Imaging (NODDI) is an important imaging technology used to evaluate the microstructure of brain tissue, which is of great significance for the discovery and treatment of various neurological diseases. Current deep learning-based methods perform parameter estimation through diffusion magnetic resonance imaging (dMRI) with a small number of diffusion gradients. These methods speed up parameter estimation and improve accuracy. However, the diffusion directions used by most existing deep learning models during testing needs to be strictly consistent with the diffusion directions during training. This results in poor generalization and robustness of deep learning models in dMRI parameter estimation. In this work, we verify for the first time that the parameter estimation performance of current mainstream methods will significantly decrease when the testing diffusion directions and the training diffusion directions are inconsistent. A robust NODDI parameter estimation method with adaptive sampling under continuous representation (RobNODDI) is proposed. Furthermore, long short-term memory (LSTM) units and fully connected layers are selected to learn continuous representation signals. To this end, we use a total of 100 subjects to conduct experiments based on the Human Connectome Project (HCP) dataset, of which 60 are used for training, 20 are used for validation, and 20 are used for testing. The test results indicate that RobNODDI improves the generalization performance and robustness of the deep learning model, enhancing the stability and flexibility of deep learning NODDI parameter estimatimation applications.

Figures (2)

• Figure 1: Overview of RobNODDI. It contains training stage and testing stage. RobNODDI performs adaptive sampling and SH fitting on the DWI patches, and then concatenates the SH coefficients into the model. Note that both adaptive sampling and SH fitting are included in the training stage, and only SH fitting is included in the testing stage.
• Figure 2: Qualitative comparison of NODDI parameter for SS and RS testing using different methods with 30 diffusion directions per shell (1000, 2000 s/mm²)