PerAF-based resting-state fMRI classifier for minimal hepatic encephalopathy.

Area of Science:

• Neuroimaging
• Medical Diagnostics
• Biomarker Discovery

Background:

• Minimal hepatic encephalopathy (MHE) is a prevalent cognitive impairment in patients with end-stage liver cirrhosis.
• Accurate diagnosis of MHE is challenging due to the lack of sensitive biomarkers and reliable diagnostic methods.
• Understanding abnormal spontaneous brain activity is crucial for MHE diagnosis.

Purpose of the Study:

• To investigate abnormal spontaneous brain activity in MHE patients.
• To evaluate the diagnostic value of four neuroimaging indicators (ALFF, fALFF, PerAF, ReHo) for MHE using Support Vector Machine (SVM).
• To identify potential biomarkers for early MHE detection.

Main Methods:

• Recruited 45 MHE patients and 40 healthy controls.
• Measured local spontaneous brain activity using Amplitude of Low Frequency Fluctuation (ALFF), Fractional Amplitude of Low Frequency Fluctuation (fALFF), Percentage Amplitude of Low Frequency Fluctuation (PerAF), and Regional Homogeneity (ReHo).
• Employed SVM analysis to build a classification model and assess diagnostic efficacy.

Main Results:

• MHE patients showed decreased ALFF in specific brain regions (AUC=0.75) and decreased ReHo (AUC=0.72).
• fALFF metrics demonstrated limited classification performance (AUC < 0.70).
• Decreased PerAF in specific frontal and parietal regions achieved high accuracy (AUC=0.83, accuracy=81.18%), outperforming other metrics.

Conclusions:

• Decreased mean PerAF in the right supramarginal gyrus, right dorsolateral superior frontal gyrus, and right middle frontal gyrus are potential neuroimaging indicators for MHE.
• These findings provide critical evidence for developing clinical screening protocols for early MHE identification.
• PerAF serves as a promising biomarker for early cognitive impairment detection in MHE patients.