The Development of Computer-Assisted Techniques for the Classification of Nerve Spike Signals

Browning, Mark Hamilton (2000) The Development of Computer-Assisted Techniques for the Classification of Nerve Spike Signals. MSc(R) thesis, University of Glasgow.

Abstract

Since the development of electronic amplification and signal recording facilities, there has been considerable interest in separating and classifying nerve spike events. Initially techniques were developed to identify spikes on the basis of amplitude but, as technology has progressed, the main interest has been in the development of techniques for classifying nerve spikes on the basis of shape. A range of strategies have been developed for performing the separation, but these strategies have been (and possibly still are) limited by the capabilities of the available hardware. These strategies and their implementations are described. A novel method for performing automated spike shape classification is described. Software has been written to implement this method, and it is applied to nerve spike data from extracellular recordings of the superficial flexor nerve of the Norway lobster (Nephrops norvegicus) and from the coxo-basal chordotonal organ and cuticular stress detector one of the crayfish (Procambarus clarkii). The results are assessed by the use of contemporaneous intracellular recordings and compared with the performance of a commercially available spike classifier, voltage thresholding techniques and an implementation of a pre-existing technique for classifying spikes. The relative merits of different strategies are considered, as well as the fundamental limitations of attempting to segregate spike data on the basis of shape alone. Technical issues relating to the implementation of a software based spike classifier are also considered.

• Item Type: Thesis (MSc(R))
• Qualification Level: Masters
• Additional Information: Adviser: Jon Barnes
• Keywords: Bioinformatics, Neurosciences
• Date of Award: 2000
• Depositing User: Enlighten Team
• Unique ID: glathesis:2000-76138
• Copyright: Copyright of this thesis is held by the author.
• Date Deposited: 19 Nov 2019 16:35
• Last Modified: 19 Nov 2019 16:35
• URI: https://theses.gla.ac.uk/id/eprint/76138