Firing rate maps of 25 hippocampal place cells simultaneously recorded in a rat running on the elevated track of a figure-8 maze, with clockwise movement in the upper loop and counterclockwise movement in the lower loop, and slowing down at the corners for food. The restricted regions with high firing rates (red color) are called the place fields. The reciprocal fields, obtained from the firing rate maps using a pseudoinverse, roughly resemble the place fields, but have negative values in surrounding regions.
True X and Y positions of the rat as compared with the positions reconstructed by different methods using the instantaneous activity of 25 place cells. The probabilistic or Bayesian methods were especially accurate, and the erratic jumps in the reconstructed trajectory were reduced by a continuity constraint using information from 2 consecutive time steps. The best error is within a factor of two of the information-theoretic limit.
All the reconstruction methods used here can be implemented as linear feedforward neural networks, in sharp contrast with the general impression that reconstruction methods such as the Bayesian method or the template matching method (direct basis)are merely mathematical techniques without any biological relevance.