Abnormal higher-order network interactions in Parkinson's disease visual hallucinations.

Area of Science:

• Neuroscience
• Systems Neuroscience
• Computational Neuroscience

Background:

• Visual hallucinations are common in Parkinson's disease (PD).
• Dysfunctional brain network communication, particularly between sensory and higher-order networks, is implicated in PD hallucinations.
• The precise functional brain alterations remain unclear.

Purpose of the Study:

• To characterize the altered functional hierarchy associated with visual hallucinations in PD using advanced dimensionality reduction techniques.
• To investigate the role of functional integration between brain networks in the pathophysiology of visual hallucinations.

Main Methods:

• Applied non-linear dimensionality reduction techniques: diffusion map embedding and t-distributed stochastic neighbour embedding (t-SNE).
• Analyzed resting-state functional MRI data from 77 individuals with PD (31 with hallucinations, 46 without) and 19 healthy controls.
• Utilized traditional functional connectivity analysis for comparison.

Main Results:

• Found compression of the unimodal-heteromodal gradient in hallucinating PD patients, indicating increased functional integration between sensory and higher-order networks.
• Observed increased connectivity between visual and default mode networks in the hallucinating group.
• t-SNE analysis revealed distinct prefrontal alterations not evident in traditional analyses.

Conclusions:

• Results suggest higher-order regions may excessively influence sensory processing in PD hallucinations.
• Abnormal brain organization is confirmed in the hallucinating phenotype of PD.
• Reduced hierarchical differentiation in sensory-perceptual systems may be a transdiagnostic vulnerability in neuropsychiatric disorders with perceptual disturbances.