Functional characterization of neurons in the early visual pathway: An exploration of spike train classification methods

• Cássio Mendes Fontes UFMG
• Ana Luiza Turchetti-Maia UFMG
• Jerome Baron UFMG
• Antônio Pádua Braga UFMG

Resumo

The understanding of how neurons interact in the visual cortex and what kinds of neurons are responsible for each interaction are still unanswered questions in the field of neurophysiology. To acquire information on such problems, sixteen neurons were mapped with visual stimulus of different sizes, contrasts but with similar patterns on their receptive fields. The data obtained was analyzed with several approaches. PCA was applied as a linear dimensionality reduction technique and the new data was separated with clustering techniques. The acquired data was also represented as the difference of time (moment in time when one spike occured minus the previous one) between spikes. Several attributes were extracted and clustered into different groups and with kernel techniques to visualize the data two large groups of neurons were identified. A support vector machine classifier (SVMC) was implemented in order to verify if an efficient classification could be performed. The main data was also separated as spikes before, during and after direct stimulation using the difference of time between spikes as data. The same attributes were extracted from these matrices and kernel techniques were applied on them to visualize the data. As before, two large groups of data were identified and isolated with clustering techniques which represented two groups of neurons.

Referências

Chen, Y. and Lin, C. (2003). Combining svms with various feature selection strategies. Technical report.

Guyon, I. and Elisseeff, A. (2003). An introduction to variable and feature selection. Journal of Machine Learning Research, 3:1157–1182.

Hartline, H. (1938). The response of single optic nerve fibers of the vertebrate eye to illumination of the retina. American Journal of Physiology.

Haykin, S. (1994). Neural Networks: A Comprehensive Foundation. Macmillan, New York.

J Baron, L. P. M. O. D. B. L. S. . N. (2007). Directional responses of visual wulst neurons to grating and plaid patterns in the awake owl. European Journal of Neuroscience.

J.R Cavanaugh, W. B. J. M. (2002). Nature and interaction of signals from the receptive field center and surround in macaque v1 neurons. Journal of Neurophysiology.

Kufler, S. (1953). Discharge patterns and functional organization of mammalian. Journal of Neurophysiology.

Peter E. Hart, R. O. D. D. G. S. (2007). Pattern Classification, chapter 9, pages 528–530. WILEY, second edition.

Queiroz, F. . B. A. P. . P. W. (2009). Sorted kernel matrices as cluster validity indexes. 2009 IFSA World Congress.

Shawe-Taylor, N. C. J. (2000). An Introduction to Support Vector Machines and other kernel-based learning methods, chapter 6, pages 93–122. Cambridge, first edition.

Smith, L. (2002). A tutorial on principal components analysis.

Wang, W. Z. D. Z. K. (2008). On kernel difference-weighted k-nearest neighbor classification. Pattern Analysis & Applications.