MIIND

MIIND simulates dynamics of neuronal populations using population density techniques to model large-scale networks of point neurons and analyze mesoscopic population activity.

Key Features:

• Model agnosticism: Agnostic to the underlying neuron model, supporting simulation of 1D and 2D dynamical systems for each population.
• Population density methods: Implements population density techniques that represent population activity geometrically and visually.
• Noise control: Provides methods for controlling noise levels within simulations.
• XML-style configuration: Networks are configured with an XML-style simulation file specifying population interactions, transmission delays, post-synaptic potentials, and desired outputs.
• Population-state visualization: Produces visual displays of each population's state during simulations.
• Data output and integration: Simulation results can be written to files or passed to Python scripts and integrated with platforms such as The Virtual Brain.
• Performance efficiency: For 1D neuron models, the population density approach demonstrates superior performance compared to direct simulation and offers advantages for 2D models.
• Scalability and insight: Uses geometric and visual descriptions of neuronal activity to aid interpretation of population dynamics and inter-population influences.

Scientific Applications:

• Multiscale bridging: Bridges microscopic (individual neurons) and mesoscopic (population networks) scales in neural modeling.
• Complex network modeling: Enables construction and analysis of complex neural population networks while reducing the complexity of managing extensive interconnections.
• Population dynamics research: Supports investigation of population-level dynamics, noise effects, transmission delays, and post-synaptic potential interactions.
• Multiscale integration: Facilitates integration into multiscale modeling workflows through output to Python and interoperability with The Virtual Brain.

Methodology:

MIIND uses population density techniques to represent populations of point neurons, providing geometric and visual representations of neuronal activity and supporting 1D and 2D dynamical system models.