In prior studies (Berger et al. 2011; Hampson et al. 2012b, Hampson et al. 2013; Opris et al. 2013), it has been shown that a multi-input multi-output model (MIMO) nonlinear dynamic model applied to spatiotemporal patterns of multiple recordings from synaptically connected neurons is capable of extracting patterns of firing related to successful task performance of this same task shown in Fig. 19.1a. It was shown that the MIMO model could be used to facilitate and recover performance when administered as patterns of electrical pulses to the same recording locations from which output signals were detected on correct trials (Hampson et al. 2012a, 2013; Berger et al. 2011). The MIMO model applied to minicolumnar data recorded by the conformal MEA probes in the animals performing the DMS task shown in Fig. 19.1 as a modeling strategy for the nonlinear dynamics underlying spike-train transformations between L2/3 and L5. This was then used to predict L5 output firing patterns from input patterns of L2/3 neural activity, as a representation of multicolumnar firing patterns (Berger et al. 2005, 2011, 2012; Hampson et al. 2012a; Song et al. 2007, 2009). In this application, the identification of spatio-temporal pattern transformation from the PFC layer 2/3 to layer 5 in MEA identified columns was formulated by the MIMO model and analyses included extraction of the first-, second- and third-order temporal firing within at least two defined minicolumns on MEAs inserted repetitively on multiple recording sessions in order to extract relevant patterns of minicolumnar activity related to successful image selection during the match phase of the task.

The same MIMO model was applied in this study to multicolumnar PFC L2/3 and L5 cell pairs recorded during the Match phase of the DMS task. Spatiotemporal patterns associated with trials of correct performance were constructed from 2–4 L2/3 and L5 confirmed MEA cell pairs recorded in five different NHPs (n = 5) performing the same task. MIMO model derived L5 output firing associated with L2/3 input firing, was used to predict performance on individual trials. Figure 19.2 shows the relationship of L5 firing to task performance with respect to the firing patterns extracted by the MIMO model. Performance of the task shown in Fig. 19.2a was differential with respect to the number of distracter images with an increase in image number associated with poorer performance. However if performance on trials was segregated and the type of MIMO output associated with performance identified, a clear distinction between the types of patterns present on successful (strong code) vs. poorer (weak code) performed trials for each number of images was clearly present as shown in Fig. 19.2b, c. The MIMO model clearly can derive appropriate firing patterns but what is more important detection of patterns that are less appropriate for successful performance can be predicted by monitoring the associated L2/3 input patterns for wrong code trials online that provides time for modifying the L5 firing pattern on the same trial.

Prior investigations applying MIMO model-derived stimulation patterns to hippocampus in the rodent provided the means to enhance performance and overcome deficits induced by pharmacological treatments (Berger et al. 2011). In this manner, predictions of the L5 output related to the successful performance were monitored online during the task to define when successful or non-successful trials were about to be completed prior to target selection in the Match phase as shown in Fig. 19.3. This online monitoring by the MIMO model provided the basis for interposing the correct (strong code) L5 firing pattern as electrical pulses delivered to the same L5 recording pads of the MEAs at the same time as normal occurrence after Match phase onset and completion of target selection (Fig. 19.3). Stimulation pulses consisted of biphasic constant current square waves, 0.5 ms per phase, adjusted in intensity (10–50 μA) to produce local field potentials monitored on adjacent L5 recoding pads on the same MEA. Model-derived stimulation was applied on 50 % of trials in each session to compare stimulated and non-stimulated trials in terms of correct performance. Effective stimulation patterns that were determined to facilitate performance under normal conditions were also applied at the same time during the trial, irrespective of prior L2/3 activity, when employed for the recovery of function as a neural prosthesis as described below (Fig. 19.5). In this manner, it was demonstrated that the MIMO model served as a prostheses for recovering decision making that required appropriate PFC columnar processing related to successful target selection in the DMS task.

The effects of MIMO generated electrical stimulation patterns averaged over all NHPs and trials (Fig. 19.4) were highly significant (p < 0.001, ANOVA). It is clear that MIMO stimulation facilitated performance above normal levels but was not consistent for the same types of trials in all animals (Fig. 19.4a, individual NHPs); however, Fig. 19.4b shows that the overall effect of MIMO stimulation was to significantly facilitate performance in all animals across trials with increased number of images (p < 0.001). Figure 19.4c shows an additional facilitatory effect of the MIMO-derived stimulation on trials with longer delays (30–60 s, 61–90 s) where performance was less impaired on stimulated versus nonstimulated trials (p < 0.001). Finally, the facilitatory effects of MIMO stimulation on target selection and execution were strongly supported by a significant decrease (F(4,401) = 9.14, p < 0.01) in the latency to make the match response (Fig. 19.4d) as a function of trial difficulty (number of images) on MIMO stimulation versus nonstimulation trials. The controls associated with these results are found in Hampson et al. (2012b).

Extensive prior investigation of features that affect cognitive processing in DMS tasks have shown that Match phase activation of PFC is altered by the modulation of dopamine influences on task-related PFC cell firing by cocaine and other agents that alter dopamine uptake (Hampson et al. 2011; Porrino et al. 2013; Arnsten and Robbins 2009). To determine similar actions on PFC columnar activity in this task firing was assessed in the same minicolumns before and after systemic injection of cocaine (0.40 mg kg⁻¹) midway through the DMS session. Figure 19.5a shows raster/PEHs for a PFC inter-laminar cell pair (L2/3 upper, L5 lower) recorded: (1) in the first 60 trials of a DMS session (IV Saline control) followed by, (2) the un-signaled administration (IV) of cocaine on trial 61 in the same session of 120 trials. It is clear that a significant reduction in match phase firing occurred in L2/3 (p < 0.001) and L5 (p < 0.001) in the cocaine versus control half of sessions (Fig. 19.5a), and that these changes in firing resembled closely the type of reduced firing that occurs on error trials (Opris et al. 2012a). The generality of this effect on match phase firing over a large population of cell pairs (n = 30) is shown in Fig. 19.5b as a significant decrease in the L2/3 cell activity (p < 0.001) relative to the saline half of the session. The reduction in L5 average firing rates in the cocaine half of the session approached but did not reach significance (p > 0.10) which perhaps reflects decreased dopamine sensitive processes (Gulledge and Jaffe 1998) relative to L2/3 cells as has been indicated in prior studies of dopamine receptor actions in these PFC layers in NHPs (Bordelon-Glausier et al. 2008; Arnsten and Robbins 2009). However, the more specific columnar firing indicator, CCHs for the single cell pair in Fig. 19.5a and for all cell pairs in Fig. 19.5b (n = 30), showed mark clear decreases in correlated firing (p < 0.001) after cocaine administration (Fig. 19.5c, d). The significance of this change with respect to cocaine’s effect on correlated firing of cells in L2/3 and L5 is shown in Fig. 19.5e as a scatter plot of correlation coefficients for control and cocaine halves of the same behavioral sessions. The lack of points along the diagonal line in the scatter plot indicates a decrease in synchronized firing in cell pairs that showed high correlations in the first half of the session, after cocaine administration in the second half of the same session. Finally, it is an important coincidence that the alterations in columnar processing produced by cocaine also influenced task performance in a manner consistent with cognitive demand as shown in Fig. 19.5f by the reduction in mean selection accuracy during the cocaine half of the session as a function of trial difficulty on trials with increased numbers of images (p < 0.001). These data are supported by other findings showing impairment in cognitive function by agents which alter activity in PFC (Arnsten 2000; Arnsten and Robbins 2009; Wang et al. 2011).