Long-term stability of visual pattern selective responses of monkey temporal lobe neurons

Abstract

Many neurons in primate inferotemporal (IT) cortex respond selectively to complex, often meaningful, stimuli such as faces and objects. An important unanswered question is whether such response selectivity, which is thought to arise from experience-dependent plasticity, is maintained from day to day, or whether the roles of individual cells are continually reassigned based on the diet of natural vision. We addressed this question using microwire electrodes that were chronically implanted in the temporal lobe of two monkeys, often allowing us to monitor activity of individual neurons across days. We found that neurons maintained their selectivity in both response magnitude and patterns of spike timing across a large set of visual images throughout periods of stable signal isolation from the same cell that sometimes exceeded two weeks. These results indicate that stimulus-selectivity of responses in IT is stable across days and weeks of visual experience.

Figure 1. Single unit responses in area TE.

Data are shown from two neurons from monkey N97. The two neurons were recorded on two different microwires (channels of electrodes bundle marked as I1 and I8) and during two different time periods. Directly below each image, the action potential responses are shown over a period of several days, with each background color corresponding to data collected from a different session. The diverse responses appear to be stable over the recording periods. At the bottom are the corresponding peristimulus time histograms for the two neurons.

Figure 2. Single unit responses in the STS.

Data are shown from two neurons collected on two different microelectrodes from monkey E98 in the same format as Figure 1 . The data were collected simultaneously for three days, after which the lower neuron was monitored for an additional 3 days.

Figure 3. Waveforms remained similar between sessions.

(a) The waveforms of the simultaneously recorded neurons shown in Figure 5 over the stable recording period. The basic shape remained similar, though the spike amplitude drifted over time. (b) Comparison in the ratio of the spike waveform amplitudes across the population (see Methods ). Stable neurons recorded across multiple sessions had amplitude ratios near 1.0, indicating that they did not change much over time. The amplitude ratio between different neurons recorded from the same session was generally much lower. (c) Correlation (r) in the depth of selectivity between the first and final sessions of all stably recorded neurons. Data shown separately for monkey N97 (open circles) and monkey E98 (black triangles) (d) Correlation (r) in the sparseness index between the first and final sessions of all stably recorded neurons.

Figure 4. Comparison of variance attributed to stimulus selectivity vs. session number across the population of stably recorded neurons (N = 72).

A 2-way ANOVA was used to compute the percent variance attributed to these two factors. (a) Distribution of variance in the spike count (40 to 440 ms following stimulus onset) over stimulus and session. (b) Same analysis as (a), but for the variance in the coefficients of the first six principal components. Principal components were computed for each neuron individually based on the mean responses to the fifteen stimuli eliciting the highest responses. For each neuron, the coefficients corresponding to each trial were then computed as the inner product of the individual trial histogram with the first six principal components (see Methods ).

Figure 5. Responses of two area TE neurons to 65 stimuli.

The two neurons were recorded simultaneously from the same electrode over a period of four days. The stimuli, presented at time = 0, are shown in the left portion of each panel. Despite the proximity of the two cells, their responses, including their basic stimulus selectivity, remained distinct over the recording sessions. Accumulated peristimulus time histograms are shown above and below the rasters of the two neurons.

Figure 6. Evaluation of stability over the population of recorded neurons.

A correlation-based similarity index (CSI) was used to characterize the similarity of the selective responses between the same neuron measured in the first and last recording sessions from the same electrode (a), between different neurons measured simultaneously on the same electrode (b), and between different neurons measured simultaneously on different electrodes (c). In the last case, only neurons for which spike waveforms were successfully maintained over at least 2 sessions were considered. Panels (d) and (e) show two neurons with the highest and lowest CSI between sessions, corresponding to the orange and purple dots in (a), respectively.