A Definition

What are neuroinformatics and computational neuroscience?

Neuroinformatics and computational neuroscience are terms that are sometimes used interchange- ably and sometimes taken to be at least partly distinct. For our purposes, we consider the terms together to encompass all computational approaches to addressing questions in neuroscience. Like the overarching field of neuroscience, it encompasses all levels of the nervous system, from molecules to behaviour, and is aimed at helping to understand the brain and ultimately improve human health.

One branch, perhaps most traditionally referred to as ”computational neuroscience” and which also encompasses ”theoretical neuroscience”, involves the development and application of mathematical models for understanding brain development, organization and function. Such models can incorporate various levels of complexity and biological realism and are used to formulate theories relating to computation and information processing at the cellular, neural network and systems levels. Thus, these models are constructed in order to gain better understanding as to how a particular function (e.g. memory) is implemented in a biological system (e.g. the rodent hippocampus). Importantly, these modeling efforts must be distinguished from the fields of neural computing or artificial intelligence. This is because biological realism is a premium for the computational neuroscientist. Thus experimentation is used to verify models, and to influence their further development.

Another major branch, most often described as ”neuroinformatics”, deals with the methods used for acquisition, analysis, and representation of data collected from neuroscience studies. This encompasses the development and application of algorithms, statistics, databases, ontologies and data standards. Sub-branches address different modalities such as neuroimaging, behavioural studies, clinical information, electrophysiology, anatomy, genomics and proteomics, and more. Importantly, many computational efforts are targeted at integration of data across studies and modalities. The rapidly increasing complexity and volume of data collected by neuroscientists has spurred the growth of research in this area.

What brings neuroinformatics and computational neuroscience together is that more and more analysis techniques incorporate and are based on formal computational models. Conversely, without neuroinformatics we cannot extract data relevant to evaluating and constraining computational models. Bringing these fields together is natural and necessary for modern neuroscience research.

Source: http://www.neuroinfocomp.ca/about