Predicting odorant quality perceptions from multidimensional scaling of olfactory bulb glomerular activity patterns.

Abstract

Odorants and their perceptions differ along multiple dimensions, requiring that a critical examination of any putative neural code directly assess the multidimensional nature of the encoding process. Previous work has examined simple, systematic odorant differences that, regardless of coding strategy, would be expected to produce simple, systematic predictions in neural and behavioral responses. In the present study, an odorant identification confusion matrix task that extracts precise quality relationships across odorants was used to determine whether spatially specific glomerular activity patterns predict perceptual quality relationships for odorants that cannot easily be classified a priori along a single chemical dimension. Multidimensional scaling (MDS) analysis of odorant pattern similarity measures derived from the comparison of [14C]-2-deoxyglucose glomerular activity pattern data yielded a two-dimensional odorant activity space that was highly significantly predictive of similarly obtained odorant perceptual spaces, uniformly across animals. These results strongly support the relevance of global spatial patterns in the olfactory bulb to the encoding of odor quality.