Identification of patients with and without minimal hepatic encephalopathy based on gray matter volumetry using a...

Area of Science:

• Neuroimaging
• Hepatology
• Machine Learning

Background:

• Minimal hepatic encephalopathy (MHE) is associated with widespread cerebral structural abnormalities, including reduced cortical thickness and altered brain parenchymal volume.
• Accurate diagnosis of MHE is crucial for managing cirrhotic patients, but traditional methods can be challenging.

Purpose of the Study:

• To evaluate the efficacy of gray matter (GM) volumetry, analyzed with a support vector machine (SVM) learning method, in distinguishing cirrhotic patients with and without MHE.
• To identify specific brain regions contributing to the differentiation between MHE and non-MHE groups.

Main Methods:

• Acquisition of high-resolution, T1-weighted magnetic resonance images from 24 cirrhotic patients with MHE and 29 cirrhotic patients without MHE (NHE).
• Voxel-based morphometry was used to assess gray matter volume (GMV) for each subject.
• An SVM classifier with leave-one-out cross-validation was employed to determine classification accuracy.

Main Results:

• The SVM algorithm utilizing GM volumetry achieved a classification accuracy of 83.02%, with a sensitivity of 83.33% and a specificity of 82.76%.
• Key discriminative GMVs were predominantly located in the bilateral frontal lobe, lentiform nucleus, thalamus, sensorimotor and visual areas, temporal lobe, cerebellum, left inferior parietal lobe, and right precuneus/posterior cingulate gyrus.

Conclusions:

• SVM analysis of GM volumetry demonstrates significant potential as a diagnostic tool for identifying MHE in cirrhotic patients.
• The identified brain regions highlight areas affected by MHE, providing insights into its neurobiological underpinnings.