Energy landscape analysis of neuroimaging data

Area of Science:

• Computational neuroscience
• Statistical physics
• Neuroimaging analysis

Background:

• Computational neuroscience models aid in understanding brain dynamics and alterations due to changing physiological conditions.
• Energy landscape analysis, a data-driven approach rooted in statistical physics (Ising model, Boltzmann machine), is reviewed for neuroimaging applications.

Purpose of the Study:

• To review and develop data-driven energy landscape analysis methods for neuroimaging data.
• To apply these methods to functional magnetic resonance imaging (fMRI) data from healthy individuals.
• To investigate the relationship between fitting accuracy, system size, and data length in neuroimaging analysis.

Main Methods:

• Review of energy landscape analysis methods, including algorithms and technical aspects.
• Application of energy landscape analysis to functional magnetic resonance imaging (fMRI) data.
• Statistical physics-based modeling, specifically the Ising model (pairwise maximum entropy model, Boltzmann machine).

Main Results:

• The study explores the application of energy landscape analysis to neuroimaging data, an area still in its early stages.
• The relationship between the accuracy of fitting, the size of the brain system analyzed, and the length of the data was inspected using fMRI data.
• The methods, while established in electrophysiology, are being adapted and evaluated for neuroimaging.

Conclusions:

• Energy landscape analysis offers a novel data-driven approach for understanding neural dynamics in neuroimaging.
• The study provides insights into the practical application and limitations of these methods with fMRI data.
• Further development and application of energy landscape analysis are warranted for advancing computational neuroscience and neuroimaging research.