Classification and characterisation of brain network changes in chronic back pain: A multicenter study.

Area of Science:

• Neuroscience
• Medical Imaging
• Computational Biology

Background:

• Chronic pain is a widespread and debilitating condition.
• The underlying neurobiological changes in the brain remain poorly understood.
• Investigating brain network alterations is crucial for understanding chronic pain.

Purpose of the Study:

• To investigate brain network architecture in chronic back pain patients.
• To identify characteristic changes in brain connectivity and network topology.
• To explore the potential of machine learning for classifying chronic pain based on brain data.

Main Methods:

• Resting-state functional magnetic resonance imaging (fMRI) data from chronic back pain patients and controls.
• Application of machine learning and deep learning (conditional variational autoencoder) for patient/control classification.
• Multislice modularity analysis to detect network reorganization patterns.

Main Results:

• Machine and deep learning models achieved significant accuracy in classifying chronic pain patients.
• Consistent disruption of network hubs, particularly in clustering coefficient and betweenness centrality, was observed.
• Evidence of modular reorganization in sensorimotor cortex, characterized by reduced pairwise links and increased connectivity with the pregenual anterior cingulate cortex.
• Intraparietal sulcus showed positive modular reorganization, suggesting a role in the chronic pain state.

Conclusions:

• Chronic pain is associated with consistent and characteristic alterations in brain network architecture.
• Sensorimotor cortex exhibits significant network reorganization, impacting its functional connectivity.
• These findings provide a deeper understanding of the central neurobiological underpinnings of chronic pain.