Computational Neuroscience Overview

Principles

Definition

Computational neuroscience is an interdisciplinary research field that employs mathematical models, theoretical analysis, and computer simulations to investigate the principles and mechanisms governing the structure and function of the nervous system. It seeks to understand how electrical and chemical signals in the brain lead to complex behaviors and cognitive abilities, providing a theoretical framework for interpreting experimental data and generating testable hypotheses.

Ontological type

Modeling Techniques

Core Theoretical Frameworks

Neuroscience Applications

Key developments

Key Figures

Population Dynamics Paradigm era

Apostolos P. Georgopoulos[1] is a central figure associated with Johns Hopkins University[3] and Karolinska Institutet[4] during this era. His Neuronal Population Coding of Movement Direction[6] established that motor cortex populations encode movement direction through distributed population codes, a foundational contribution that shaped how researchers approached population dynamics and ensemble coding in perception, memory, and motor systems. Andrew B. Schwartz[2] is similarly linked to Johns Hopkins University[3] and the University of Minnesota[5] during this era. In the same Neuronal Population Coding of Movement Direction[6], Schwartz helped articulate how population-level representations support motor control and ensemble coding, a cornerstone concept that spurred electrophysiological signal processing and modeling within this era.

Standardized Connectomic Frameworks era

Scott Makeig [1], associated with University of California, San Diego [3] and the University of California System [4], emerges as a leading figure in this era. Makeig [1] contributed to EEGLAB [7], an open source toolbox for analysis of single-trial EEG dynamics including independent component analysis, enabling standardized preprocessing and reproducible analyses across labs. Stephen M. Smith [2], affiliated with Boston University [5] and University College London [6], contributed to A global optimisation method for robust affine registration of brain images [8] and Consistent resting-state networks across healthy subjects [9], advancing cross-subject alignment and reproducible functional networks. These works reinforced cross-subject comparability and reproducibility in large-scale connectomics, aligning with standardized atlas-based pipelines and graph-theoretic analyses that defined this era.

Deep Graph Integration era

Wenming Zheng [1] is associated with Southeast University [3] and University of Macau [4] during the Deep Graph Integration era. His key contribution, EEG Emotion Recognition Using Dynamical Graph Convolutional Neural Networks [7], demonstrated dynamic graph-based processing of EEG data to classify emotions, a significant step in applying graph neural networks to brain-state decoding in this era. Steven P. Miller [2] is affiliated with University of California, San Francisco [5] and Johns Hopkins University [6] during the Deep Graph Integration era. His notable contribution, BrainNetCNN: Convolutional neural networks for brain networks; towards predicting neurodevelopment [8], introduced brain network CNNs to predict neurodevelopmental trajectories from connectome-like data, catalyzing graph-based deep learning approaches for brain network analysis in this era.