Multiple electrode analysis of the synchronous spatiotemporal patterns of Purkinje cell complex spike activity in mammalian cerebellum 小脳におけるプルキンエ細胞複雑スパイク活動の同期性時空間パターンの多数電極による解析

博士論文

application/pdf

Complex spike activity In Purkinje cells in rat cerebellar cortex was studied utilizing multiple electrode recording techniques. Two approaches were used to analyze spatiotemporal relationships within the complex data arrays recorded during these experiments: a) grouping cells with similar spatiotemporal firing patterns via the projection method and b) so-called principal component analysis. The grouping technique enabled the visualization of relationships within an entire neuron set, through the projection of cross-correlation vectors from hyperdimensional to lower dimensional space. Each cluster of cross-correlation vectors corresponded well to the rostrocaudal organization of climbing fiber activity in cerebellar cortex. Application of principal component multivariate analysis revealed major components of complex spatiotemporal variance in climbing fiber activity of multiple Purkinje cells. A maximum variance of 30% may be ascribed to the first and second components, which corresponded, respectively, to synchronicity and spatial grouping In the spatiotemporal organization. These analyses permitted a global and quantitative description of the simultaneous activity of groups of individual neurons. Previous multi-electrode studies of cerebellar Purkinje cell activity analyzed temporal interactions between different folia oriented in a rostrocaudal band and between folia tn different parts of the rat cerebellar cortex^<18>. As the number of cells recorded increases from 32 in our first studies^<16> to 96^<18>, more sophisticated methods of data analysis have been required. Recently, multichannel neuronal activity has been recorded using optical measurement s6 as well as multiple microelectrode^<1,9>. However, methods for the analysis of such complex multiple neuronal activity is still lacking^<5,9>. This paper details two techniques: a) projection method and b) principal component analysis. The latter is often applied to analyze the magnitude of responses to various stimuli^<2,3,8>. Preliminary reports have been presented^<4,17>.

Article

富山医科薬科大学・博士(医学)・乙第190号・福田正治・1993/1/13