Position reconstruction of rodent movements based on neural spike information of place cells

Place cells are located in the hippocampus of the brain and play an important role for spatial navigation. In this study neural spike activity of freely moving rats along with the position of the rats was acquired. The study was performed to investigate if position reconstruction is possible if the rat is freely moving in open arenas with and without barriers of different rectangular sizes based on neural recordings. The neural spike activity was measured using tetrodes from 6 chronically implanted rats in CA1 and subiculum regions. The position of the rats was monitored using a video tracking system. In the encoding step spike activity features of the place cells and position information from the rats were used to train a computer system. In the decoding step, the position was estimated from the neural spiking activity. Different reconstruction methods were implemented: (i) Bayesian 1-step and (ii) Bayesian 2-step. The results show, that the reconstruction methods are able to determine the position of the rat in this 2-dimensional rectangular arenas from the cell activity. Better accuracies were achieved for CA1 regions because the firing fields show more localized spots. Furthermore, higher accuracies are achieved with a higher number of place cells and higher firing rates.