Optimisation and impact of gradient waveform modulation on Non-uniform Oscillating Gradient Spin-Echo sequences for microstructural characterisation

Neurological diseases often result in changes at microscopic scales in the nervous system, emphasising the need for non-invasive imaging techniques that can quantify these alterations as potential biomarkers for diagnosis. Diffusion-weighted magnetic resonance imaging (DWI), particularly using modulated gradient spin-echo (MGSE) sequences, has significantly advanced in revealing tissue microstructure by probing molecular diffusion. Among the MGSE sequences, the Non-uniform Oscillating Gradient Spin-Echo (NOGSE) sequence generates a contrast based on selective microstructure sizes through a signal decay-shift, rather than probing conventional decay rates. In this study, we evaluate the performance of NOGSE in estimating microstructure sizes using a preclinical MRI scanner. Our results show that while sharp, instantaneous gradient modulations maximise the decay-shift, smooth gradient modulations still provide meaningful contrast. Through a combination of phantom experiments, numerical simulations and information-theoretic analysis, we optimise NOGSE parameters for accurate microstructural size estimation under both sharp and smooth gradient modulations. We identify optimal NOGSE parameters that are compatible with preclinical hardware constraints, providing reliable microstructure size characterisation. Especially smooth gradient modulations expand the range of compatible MRI scanners and are almost suitable for in-vivo applications. These findings represent a step toward developing quantitative imaging tools for probing microstructural features in diffusion-weighted magnetic resonance imaging.