# Gray matter microstructural alterations and their correlation with systemic biomarkers in hepatic encephalopathy: a NODDI study using gray-matter based spatial statistics

## TL;DR

This study uses advanced MRI techniques to detect brain microstructural changes in hepatic encephalopathy and links them to blood markers.

## Contribution

The study introduces NODDI and GBSS to explore microstructural brain changes in HE and their correlation with systemic biomarkers.

## Key findings

- HE patients showed decreased neurite density in multiple brain regions and increased orientation dispersion in the cerebellum.
- NDI in the globus pallidus correlated with bilirubin and prothrombin levels, while ODI correlated with hemoglobin.
- Findings suggest a potential 'double-hit' model linking toxic metabolites and systemic factors to brain changes in HE.

## Abstract

Hepatic encephalopathy (HE) involves complex neurobiological changes that are often difficult to quantify using conventional MRI. This study aims to utilize Neurite Orientation Dispersion and Density Imaging (NODDI) combined with Gray-matter Based Spatial Statistics (GBSS) to characterize microstructural alterations in patients with HE and explore their relationship with clinical biochemical markers, specifically within the globus pallidus (GP).

Thirty-three patients with HE and 31 healthy controls underwent 3 T MRI including a multi-shell diffusion protocol for NODDI. GBSS was performed to assess differences in the Neurite Density Index (NDI) and Orientation Dispersion Index (ODI). Pearson correlation analyzed relationships between GP NODDI parameters and blood biochemical indices.

HE patients exhibited significantly decreased NDI across widespread cortical and sub-cortical regions (frontal, parietal, temporal, cingulate, insula, thalamus) and increased ODI in the posterior cerebellum/vermis. Exploratory ROI analysis of the globus pallidus (GP)—a region known for manganese deposition but showing no significant group-level differences in this study- revealed that, the NDI of the right GP showed positive correlations with indirect bilirubin and prothrombin international normalized ratio (all uncorrected p < 0.05), while the ODI of the left GP positively correlated with hemoglobin concentration (uncorrected p = 0.046).

NODDI reveals extensive microstructural alterations consistent with reduced neurite density index and cerebellar disorganization in HE. The dissociated correlation patterns of GP NDI and ODI with distinct blood markers may be compatible with a hypothetical “double-hit” pathophysiological model: toxic metabolite accumulation may drive cellular swelling (increased NDI), while systemic factors like anemia may reduce structural complexity (decreased ODI). However, these exploratory associations do not allow causal inference. These findings highlight NODDI could be a useful tool for monitoring the progression and metabolic impact of HE.