Back to search

Aristotle University Of Thessaloniki (AUTH)

Ανάπτυξη και εφαρμογή αλγορίθμων νευροπληροφορικής για την ανάλυση βιοηλεκτρικών σημάτων

Abstract

dc:description

Spike sorting procedures are based on the assumption that all the action potential traces of a particular neuron have nearly the same amplitude and shape. The shapes of recorded spike waveforms mainly depend on neuron’s geometry and on its distance to the recording electrode. The goal of a spike sorting routine is to process and analyze the usually composite recorded signals in order to identify the number of active neurons and extract detailed time courses of their spiking activity. Related algorithms constitute the core methodological component in Neuroinformatics, ranging from traditional neurophysiological experiments and clinical/neuroscience studies to cortex-machine interfaces. In this thesis various algorithmic procedures are proposed to successfully address the open challenges in the spike sorting domain. As a first step, a widely used spike sorting algorithm is being evaluated for single-electrode nerve trunk recordings. The algorithm is based on Principal Component Analysis (PCA) for spike feature extraction. In the neuroscience literature it is generally assumed that the use of the first two or most commonly three principal components is sufficient. We estimate the optimum PCA-based feature space by evaluating the algorithm’s performance on simulated series of action potentials. Correlated and white Gaussian noise are superimposed to account for biological and artificial jitter in the recordings. We introduce a new metric to define clustering error considering over-clustering more favorable than under-clustering, as proposed by experimentalists. We report that the employment of more than three principal components is in general beneficial in every case considered. Furthermore, a number of modifications are made to the open source nev2lkit software to enable systematic investigation of the parameter space. The enhanced version of the software is available online. Next, the challenges of background noise and spike overlaps, which pose problems in contemporary spike sorting strategies, are addressed. A hybrid scheme is introduced, that combines the robust representation of spike waveforms to facilitate the reliable identification of contributing neurons with efficient data learning to enable the precise decomposition of coactivations. The introduced algorithm is better characterized as semi-supervised, noise-assisted strategy of empirical nature. Efficiency is inherited from the incorporation of well-established algorithms. Moreover, robustness is guaranteed by adaptation to the actual noise properties of a given dataset. Its validity has been verified via extensive experimentation, using realistically simulated data, under different levels of noise. Finally, a graph-theoretical algorithmic procedure is proposed that successfully resolves the identification of the actual number of active neurons in a neural recording. This decision is still an open issue in spike sorting, with sparsely firing neurons and background activity the most influencing factors. Dimensionality reduction coupled with a modern, efficient and progressively-executable clustering routine proved to achieve higher performance standards than popular spike sorting methods. The proposed method is validated extensively using simulated data for different levels of background noise and found to outperform conventional techniques even in the case of a sparse-firing neuron

Degree

thesis:*
Grantor dc:publisher
Aristotle University Of Thessaloniki (AUTH)
Year dc:date
2010

Author and committee

dc:creator, dc:contributor.*
Authors dc:creator
  • Adamos, Dimitrios
  • Αδάμος, Δημήτριος

Subjects

dc:subject × 20

Rights

Language dc:language
gre

Identifiers

dc:identifier.*
Identifier
10.12681/eadd/24370

Chain of custody

source
Harvested from
Greek National Archive of PhD Theses
Base URL
phdtheses.ekt.gr/eadd_oai/request
Last updated
2026-11-18
Source record
OAI-PMH GetRecord
citation

Adamos, Dimitrios; Αδάμος, Δημήτριος. Ανάπτυξη και εφαρμογή αλγορίθμων νευροπληροφορικής για την ανάλυση βιοηλεκτρικών σημάτων. Aristotle University Of Thessaloniki (AUTH), 2010. http://hdl.handle.net/10442/hedi/24370