Identification of minimal hepatic encephalopathy based on dynamic functional connectivity

Area of Science:

• Neuroscience
• Medical Imaging
• Hepatology

Background:

• Minimal hepatic encephalopathy (MHE) is a subtle neurological complication in cirrhosis patients.
• Current diagnostic methods may not fully capture the neural dysfunction associated with MHE.
• Dynamic functional connectivity (DFC) offers a novel approach to assess brain activity over time.

Purpose of the Study:

• To evaluate the efficacy of DFC metrics in differentiating MHE patients from cirrhotic patients without hepatic encephalopathy (noHE) and healthy controls (HCs).
• To compare the classification performance of DFC metrics against static functional connectivity (SFC) metrics.
• To identify specific brain regions that serve as discriminative features for MHE detection using DFC.

Main Methods:

• Resting-state functional MRI data were acquired from 62 cirrhosis patients (30 MHE, 32 noHE) and 41 HCs.
• The sliding time window approach was employed to calculate DFC characteristics: strength, stability, and variability.
• A linear support vector machine with leave-one-out cross-validation was used for classification, comparing DFC and SFC features.

Main Results:

• DFC strength (DFC-Dstrength) demonstrated superior classification accuracy: 72.5% (MHE vs. noHE), 84% (MHE vs. HCs), and 88% (noHE vs. HCs).
• DFC-Dstrength outperformed SFC, yielding accuracy improvements of 10.5% (MHE vs. noHE), 8% (MHE vs. HC), and 14% (noHE vs. HCs).
• Seven brain nodes, including the left inferior parietal lobule and right insula, were identified as key discriminators for MHE.

Conclusions:

• DFC metrics, particularly DFC-Dstrength, offer enhanced classification accuracy for identifying MHE in cirrhosis patients.
• Dynamic functional connectivity analysis is valuable for capturing neural processes and identifying disease-related biomarkers for MHE.
• These findings support the potential of DFC as a sensitive tool for early MHE detection and management.