Thermal noise lowers the accuracy of rotationally invariant harmonics of diffusion MRI data and their robustness to experimental variations.

Purpose: Rotational invariants (RIs) are at the root of many dMRI applications. Among others, they are presented as a sensible way of reducing the dimensionality of biophysical models. While thermal noise impact on diffusion metrics has been well studied, little is known on its effect on spherical harmonics-based RI (RISH) features and derived markers. In this work, we evaluate the effect of noise on RISH features and downstream Standard Model Imaging (SMI) estimates.

Theory and methods: Using simulated and test/retest multishell MRI data, we assess the accuracy and precision of RISH features and SMI parameters in the presence of thermal noise, as well as its robustness to variations in protocol design. We further propose and evaluate correction strategies that bypass the need of rotational invariant features as an intermediate step.

Results: Both RISH features and SMI estimates are impacted by SNR-dependent Rician biases. However, higher-order RISH features are susceptible to a secondary noise-related source of bias, which not only depends on SNR, but also protocol and underlying microstructure. Rician bias-correcting techniques are insufficient to maximize the accuracy of RISH and SMI features, or to ensure consistency across protocols. SMI estimators that avoid RISH features by fitting the model to the directional diffusion MRI data outperform RISH-based approaches in accuracy, repeatability, and reproducibility across acquisition protocols.

Conclusions: RISH features are increasingly used in dMRI analysis, yet they are prone to various sources of noise that lower their accuracy and reproducibility. Understanding the impact of noise and mitigating such biases is critical to maximize the validity, repeatability, and reproducibility of dMRI studies.