Figure 7.

The impact of NE on cortical associative memory performance. Simulations presented here show recall performance when odors are distorted or perturbed with respect to the learned representation. Degree of perturbation is indicated by the normalized Euclidean distance between the recalled and the learned odor (x-axis). (A) Pyr firing rates as a function of degree of perturbation and NE levels during learning. (B) Cortical detection levels as a function of degree of perturbation and NE levels during learning. (C) Sparseness of learned synaptic weights as a function of learning and NE levels. (D) Behavioral results showing that reward-driven odor detection has substantially lower detection thresholds (D_ii) than spontaneous detection (Di) and that the difference in thresholds depends on functioning NE receptors [reprinted from Escanilla et al. (2012)]. The graphs in (D_i) show average investigation times in response to three presentations of mineral oil (MO) and one presentation of odorant (O1) at 10⁻⁶, 10⁻⁴, and 10⁻² Pa. Note that rats only detected 10⁻² Pa odorants, as indicated by a significant increase in investigation time. (D_ii) The graph shows the percent correct choice made during 20 trials during which a diluted odorant was rewarded in the presence of a second pot containing unrewarded MO. Note that rats were able to detect odorants as low as 10⁻⁶ Pa when encouraged to do so with a food reward. (D_iii) Detection of low concentration odorants in the forced choice task was impaired at 10⁻⁶ and 10⁻⁴ Pa when NE receptors were blocked during the behavioral task.