Neural network based pattern matching and spike detection tools and services— in the CARMEN neuroinformatics project

Citation

Fletcher M, Liang B, Smith L, Knowles A, Jackson T, Jessop M & Austin J (2008) Neural network based pattern matching and spike detection tools and services— in the CARMEN neuroinformatics project. Neural Networks, 21 (8), pp. 1076-1084. http://www.sciencedirect.com/science/journal/08936080; https://doi.org/10.1016/j.neunet.2008.06.009

Abstract
In the study of information flow in the brain, component processes can be investigated using a range of electrophysiological and imaging techniques. Although data is difficult and expensive to produce, it is rarely shared and collaboratively exploited. The Code Analysis, Repository and Modelling for e-Neuroscience (CARMEN) project addresses this challenge through the provision of a virtual neuroscience laboratory: an infrastructure for sharing data, tools and services. Central to the CARMEN concept are distributed CARMEN Active Information Repository Nodes (CAIRNs) which provide: data and metadata storage, new and third party / legacy services and tools. In this paper, we describe the CARMEN project as well as the CAIRN infrastructure. In particular, we will introduce an adapted version of the Signal Data Explorer (SDE) tool and a discussion of spike detection services. The SDE tool provides data visualization, signal processing, pattern matching and act as a client for the CAIRN Grid services. The SDE was developed for use in the aero-engine domain but is being expanded and used in CARMEN to visualise and search raw and derived neuroscience data. The SDE uses an AURA (Advanced Uncertain Reasoning Architecture) neural network to perform extremely fast pattern matching. The papers shows SDE and the AURA pattern match services detecting spikes from neuronal data and how the tools can be used to search for complex conditions comprising of many different patterns across the large data sets that are typical in neuroinformatics. The SDE provides a valuable means of querying patterns from the CARMEN data depository which is complementary to conventional text based querying (using metadata, etc.) to provide further insight and annotations where possible. Spike detection services which use wavelet and morphology techniques are discussed and have been shown to outperform traditional thresholding and template based systems. A number of different spike detection and sorting techniques will be supplied to users of the CARMEN infrastructure to allow users to compare the performance.

Keywords
Neuroinformatics; Pattern matching; Spike detection

Journal
Neural Networks: Volume 21, Issue 8