Acoustic signals are represented by the joint response of neural populations along the auditory pathway. As much as traditional signal processing tools-PST and interval histograms among many-have been used to study discharge patterns, they do not elucidate what stimulus features are represented and the fidelity of that coding. Information theory provides not only a theoretic framework for signal coding and decoding, but also a measurement tool, with distance measures between responses to stimulus pairs. In measuring Kullback-Leibler distances, we have quantified neural coding in discharge patterns of lateral superior olive neurons and found the coding to be quite complex. Population simulation studies demonstrate that response unit variety can enhance coding efficiency.

1998-1999 Mathematics in Biology