Article Figures & Data
Figures
- Figure 1.
Histologic reconstruction of recording sites. A, Schematic from a rat brain atlas (Paxinos and Watson, 1986) indicating the general region of the posterior piriform cortex Layers I, II, and III in a coronal view (1.4 mm posterior to bregma). B, C, Coronal sections taken from two rat brains showing DAPI (blue) and DiI (pink) staining of nuclei and electrode tracts, respectively. Scale bar indicates 0.75 mm. Electrode tips are indicated by arrowheads. Animal in B was implanted with a microwire array; animal in C with a silicon probe.
- Figure 2.
Odor-taste association paradigm. A, Sequence of procedures in the experimental paradigm. B, Preferences for the saccharin-paired odor (odor B) obtained before and after conditioning for each animal (gray lines). Averages (±SEM) over animals (n = 29) are shown in color. C, Change in preference (post–pre conditioning), averaged (±SEM) over animals.
- Figure 3.
Responses to intraoral odor solutions recorded from exemplar pPC neurons. A–D, Top panels show 100 randomly selected waveforms (gray) in the single neuron cluster; middle panels are spike raster plots showing all action potentials for all trials aligned on stimulus delivery (t = 0); and bottom panels show average firing rate (±SEM) over trials in response to odors A and B. Rate plots are calculated using a 500-ms sliding window to illustrate the temporal profile of responses, but lack the temporal resolution of raster plots because of the size of the smoothing window. Neurons in A, B were recorded before conditioning; neurons in C, D after conditioning. Dashed lines indicate offset of “early” and “late” response epochs.
- Figure 4.
Odor responsiveness in the population of pPC neurons. A, B, Proportion of pPC neurons exhibiting a significant response relative to baseline, before (A) and after (B) conditioning. Responsiveness was calculated using a 500-ms sliding window. Dashed lines indicate offset of “early” and “late” response epochs. No significant differences in responsiveness were observed (χ2 test comparing proportions between groups).
- Figure 5.
Discriminability between odors A, B at the single neuron level. A, Effect size (Cohen’s d) of the difference in the response to odors A and B, averaged (±SEM) over all neurons recorded before and after conditioning (n = 152 and 147, respectively), during baseline, and early and late stimulus response epochs. B, Cohen’s d obtained from the early and late response epochs for all odor-responsive neuron (n = 36 before and after conditioning). Letters correspond to the example responses in Figure 3.
- Figure 6.
Single-trial decoding of odor identity from ensemble responses. A, Discriminability between odors A and B, averaged (±SEM) over all sessions with n > 1 neurons before and after conditioning (n = 20), during baseline, and early and late stimulus response epochs. Bi, Discriminability as a function of time for an example ensemble recorded after conditioning (n = 10 neurons). Stimulus period (2500 ms following stimulus onset) is highlighted. Bii, Firing rate patterns for all neurons in the example ensemble shown in Bi. Responses are shown in a series of sliding windows (500-ms window size, 100-ms step size). C, Discriminability of ensemble responses to the same odor in different response epochs. *t test: p < 0.05.
- Figure 7.
Relation between ensemble decoding accuracy and odor palatability. A, Ensemble discriminability between odors A and B during the late response epoch as a function of relative preference for odor B for each animal (n = 19). No significant relation was detected by linear regression. B, Ensemble discriminability between odors A and B during the late response epoch before and after conditioning for each animal (n = 19). *Regression: p < 0.05.
In this issue
