Construction of a neural network model for the prefrontal cortex based on experimental data using AdEx neurons

• Grant number: 24/16557-9
• Support Opportunities: Scholarships in Brazil - Master
• Start date: February 01, 2025
• End date: January 31, 2027
• Field of knowledge: Biological Sciences - Biophysics - Biophysics of Processes and Systems
• Principal Investigator: Antonio Carlos Roque da Silva Filho
• Grantee: Pedro Ribeiro Pinheiro
• Host Institution: Faculdade de Filosofia, Ciências e Letras de Ribeirão Preto (FFCLRP). Universidade de São Paulo (USP). Ribeirão Preto , SP, Brazil

Abstract

The prefrontal cortex (PFC) is considered to be one of the most studied brain regions in recent years, being related to important executive phenomena such as decision-making, conscious perception and working memory. However, there are still few initiatives that aim to integrate the growing amount of experimental data into a coherent computational model capable of reproducing in vivo behavior. A recent study made a major advance towards this goal, using a simplified version of the AdEx (adaptive exponential integrate-and-fire) neuron model, called simpAdEx, together with physiological and anatomical data to build a computational model of the PFC capable of simulating the so-called asynchronous irregular (AI) state of activity characteristic of wakefulness. However, because it presents fewer parameters than the original model, the simpAdEx model is limited in its ability to reproduce neuronal behaviors usually related to different cognitive states. This work aims to expand the original model by using the complete AdEx model, aiming to simulate not only the AI ¿¿state, but also synchronous oscillatory regimes, which will later be characterized in light of the relevant literature. Another intended contribution of this project is the implementation of the PFC model in the NEST neurosimulator, unlike the original model that was implemented in Matlab. NEST is a free and open-source program that was created specifically for building models in computational neuroscience. It has its own domain-specific language and a wide range of tools and resources for simulating neural networks. This initiative aims to make the model available on an open platform accessible to the computational neuroscience community, aiming at higher standards of efficiency, quality and replicability.