Brain entropy changes in classical trigeminal neuralgia

Area of Science:

• Neuroscience
• Medical Imaging
• Neurology

Background:

• Classical trigeminal neuralgia (CTN) is a debilitating chronic neuropathic facial pain disorder with incompletely understood pathological mechanisms.
• Resting-state functional magnetic resonance imaging (rs-fMRI) offers insights into functional brain changes in CTN.
• The precise spatial distribution of neural complexity alterations in CTN remains unclear.

Purpose of the Study:

• To investigate the spatial distribution of brain entropy (BEN) alterations in CTN patients.
• To explore the relationship between BEN changes and clinical variables in CTN.
• To evaluate the efficacy of machine learning in classifying CTN patients based on BEN patterns.

Main Methods:

• rs-fMRI data from 85 CTN patients and 79 healthy controls (HCs) were analyzed.
• Brain entropy (BEN) was calculated to assess neural complexity.
• Sixteen machine learning algorithms were employed for classification, with the best-performing model selected.

Main Results:

• CTN patients showed increased BEN in the thalamus and brainstem, and decreased BEN in the inferior semilunar lobule compared to HCs.
• A low positive correlation was observed between thalamic BEN values and neuropsychological assessments.
• The Conditional Mutual Information Maximization-Random Forest (CMIM-RF) model achieved the highest classification accuracy (AUC = 0.801).

Conclusions:

• BEN alterations in the thalamus, pons, and inferior semilunar lobule are associated with CTN.
• Machine learning effectively classifies CTN patients from HCs using BEN data.
• These findings offer novel insights into CTN neuropathology and potential diagnostic/therapeutic targets.