Algorithm Designs And Applications For The Quantitative Analysis Of Neuronal Action Potentials In Hippocampus

Action potentials (AP) are considered as carriers of signal transmission and signal processing in the brain. Recording and analysis of APs are one of the major means for the investigations of neuronal mechanisms. The rapid developments of neural signal recording techniques result a large amount of experimental data. Quantitative analyses of action potentials' indexes by using manual methods are tedious and fallible. Because of the varieties in the experimental recordings and in the experimental goals, there is a lack of commercial software for the automatic analysis. Therefore, developments of new automatic algorithms for the quantitative analyses of AP experimental data are of importance to advance the investigations in the area of neural electrophysiology.In the present the thesis, the automatic algorithms were investigated for the quantitative analyses of extracellular APs in the rat hippocampus in vivo, including the evoked population spike (PS) of electric stimulation and spontaneous unit activity. (1) For the amplitude analysis of evoked PS, a novel second-order derivative detection algorithm was developed. The algorithm was able to detect every spikes in the PS responses with multiple spikes. In addition, another algorithm called coastline burst index was proposed to calculate the PS's amplitudes. (2) A new threshold algorithm was designed to extract unit spikes. The thresholds for the detection of unit signals were determined by using the recordings without spontaneous unit signals during the inhibitory periods following electric stimulations. It enhanced the precise of the threshold algorithm. (3) By using the new algorithm, we investigated the correlations between the unit discharges in the pyramidal layer and theta rhythms in the apical dendrite layer in hippocampal CA1 region. The effects of high potassium concentration on these correlations were also investigated.The result showed that (1) the precise of second-order derivative detection algorithm was better than the common used extremum search algorithm. The algorithm was especially useful for the analysis of PSs with multiple spikes. The precise of coastline burst index algorithm was similar to the extremum search algorithm, but was of much less time consumption. (2) The new threshold algorithm for the unit spike detection was simply realized with little time consumption. Because its threshold was determined by pure noise signal recordings, the results avoided the influence from the spontaneous unit spike. (3) There were more unit neuronal firings during the positive phase of dendrite theta rhythms. The firing rate was highest near the positive peak of theta wave and decreased gradually from the positive peak to the negative peak of the dendrite theta rhythms. This difference became more obvious following the increase of extracellular concentration of potassium.The algorithms developed in this thesis for the automatic analysis of evoked PS potentials and spontaneous unit spikes are simple and precise. Their uses in analyzing experimental data can enhance the efficiency of research work. In addition, the results involved the correlations between the unit discharges and the theta rhythms in the hippocampal CA1 region are of significance for the investigation of mechanisms of neuronal plasticity.