Computational neurostimulation for Parkinson's disease.

Area of Science:

• Computational neuroscience
• Neuroimaging and stimulation techniques
• Neurological and psychiatric disorders

Background:

• Deep brain stimulation (DBS) is effective for neurological and psychiatric conditions but has suboptimal efficacy and side effects.
• The mechanisms underlying DBS effects, both beneficial and detrimental, are not fully understood, hindering therapeutic advancements.
• Current biophysical models offer insights into local and network-level neurophysiological changes but often fail to predict behavioral outcomes.

Purpose of the Study:

• To advocate for computational models that link DBS to specific network computations and behavior.
• To propose mesoscopic models focusing on oscillations within cortico-basal-ganglia networks as crucial for understanding DBS.
• To highlight the potential of computational neurostimulation for advancing DBS therapies.

Main Methods:

• Review of existing biophysical modeling approaches for DBS.
• Argument for a shift towards computational models focusing on network computations and behavior.
• Proposal for mesoscopic models emphasizing oscillations in cortico-basal-ganglia networks.

Main Results:

• Biophysical models explain neurophysiological changes but not behavioral consequences of DBS.
• Computational models focusing on network computations are essential for predicting DBS effects on behavior.
• Mesoscopic models of oscillations offer a promising avenue for understanding DBS in Parkinson's disease.

Conclusions:

• Further progress in DBS requires models that bridge neurophysiology and behavior.
• Computational neurostimulation, particularly mesoscopic models of oscillations, is key to optimizing DBS therapies.
• This approach promises a new era of rationally designed DBS advancements for neurological and psychiatric disorders.