Bagrat Amirikian

Mapping of the preferred direction in the motor cortex

Directional tuning is a basic functional property of cell activity in the motor cortex. Previous work has indicated that cells with similar preferred directions are organized in columns perpendicular to the cortical surface. Here we show that these columns are organized in an orderly fashion in the tangential dimension on the cortical surface. Based on a large number of microelectrode penetrations and systematic exploration of the proximal arm area of the motor cortex while monkeys made free reaching 3D movements, it was estimated that (i) directional minicolumns are ≈30 μm in width, (ii) minicolumns with similar preferred directions tend to occur in doublets or triplets, and (iii) such minicolumns tend to repeat every ≈240 μm (estimated width of a column), with intermediate preferred directions represented in a gradient. These findings provide evidence for an orderly mapping of the preferred direction in the motor cortex.

Modular organization of directionally tuned cells in the motor cortex: Is there a short-range order?

We investigated the presence of short-range order (<600 μm) in the directional properties of neurons in the motor cortex of the monkey. For that purpose, we developed a quantitative method for the detection of functional cortical modules and used it to examine such potential modules formed by directionally tuned cells. In the functional domain, we labeled each cell by its preferred direction (PD) vector in 3D movement space; in the spatial domain, we used the position of the tip of the recording microelectrode as the cells coordinate. The images produced by this method represented two orthogonal dimensions in the cortex; one was parallel (“horizontal”) and the other perpendicular (“vertical”) to the cortical layers. The distribution of directionally tuned cells in these dimensions was nonuniform and highly structured. Specifically, cells with similar PDs tended to segregate into vertically oriented minicolumns 50–100 μm wide and at least 500 μm high. Such minicolumns aggregated across the horizontal dimension in a secondary structure of higher order. In this structure, minicolumns with similar PDs were ≈200 μm apart and were interleaved with minicolumns representing nearly orthogonal PDs; in addition, nonoverlapping columns representing nearly opposite PDs were ≈350 μm apart.

Functional Magnetic Resonance Imaging of Visual Object Construction and Shape Discrimination: Relations among Task, Hemispheric Lateralization, and Gender

We studied the brain activation patterns in two visual image processing tasks requiring judgements on object construction (FIT task) or object sameness (SAME task). Eight right-handed healthy human subjects (four women and four men) performed the two tasks in a randomized block design while 5-mm, multislice functional images of the whole brain were acquired using a 4-tesla system using blood oxygenation dependent (BOLD) activation. Pairs of objects were picked randomly from a set of 25 oriented fragments of a square and presented to the subjects approximately every 5 sec. In the FIT task, subjects had to indicate, by pushing one of two buttons, whether the two fragments could match to form a perfect square, whereas in the SAME task they had to decide whether they were the same or not. In a control task, preceding and following each of the two tasks above, a single square was presented at the same rate and subjects pushed any of the two keys at random. Functional activation maps were constructed based on a combination of conservative criteria. The areas with activated pixels were identified using Talairach coordinates and anatomical landmarks, and the number of activated pixels was determined for each area. Altogether, 379 pixels were activated. The counts of activated pixels did not differ significantly between the two tasks or between the two genders. However, there were significantly more activated pixels in the left (n = 218) than the right side of the brain (n = 161). Of the 379 activated pixels, 371 were located in the cerebral cortex. The Talairach coordinates of these pixels were analyzed with respect to their overall distribution in the two tasks. These distributions differed significantly between the two tasks. With respect to individual dimensions, the two tasks differed significantly in the anterior-posterior and superior-inferior distributions but not in the left-right (including mediolateral, within the left or right side) distribution. Specifically, the FIT distribution was, overall, more anterior and inferior than that of the SAME task. A detailed analysis of the counts and spatial distributions of activated pixels was carried out for 15 brain areas (all in the cerebral cortex) in which a consistent activation (in 3 subjects) was observed (n = 323 activated pixels). We found the following. Except for the inferior temporal gyrus, which was activated exclusively in the FIT task, all other areas showed activation in both tasks but to different extents. Based on the extent of activation, areas fell within two distinct groups (FIT or SAME) depending on which pixel count (i.e., FIT or SAME) was greater. The FIT group consisted of the following areas, in decreasing FIT/SAME order (brackets indicate ties): GTi, GTs, GC, GFi, GFd, [GTm, GF], GO. The SAME group consisted of the following areas, in decreasing SAME/FIT order: GOi, LPs, Sca, GPrC, GPoC, [GFs, GFm]. These results indicate that there are distributed, graded, and partially overlapping patterns of activation during performance of the two tasks. We attribute these overlapping patterns of activation to the engagement of partially shared processes. Activated pixels clustered to three types of clusters: FIT-only (111 pixels), SAME-only (97 pixels), and FIT + SAME (115 pixels). Pixels contained in FIT-only and SAME-only clusters were distributed approximately equally between the left and right hemispheres, whereas pixels in the SAME + FIT clusters were located mostly in the left hemisphere. With respect to gender, the left-right distribution of activated pixels was very similar in women and men for the SAME-only and FIT + SAME clusters but differed for the FIT-only case in which there was a prominent left side preponderance for women, in contrast to a right side preponderance for men. We conclude that (a) cortical mechanisms common for processing visual object construction and discrimination involve mostly the left hemisphere, (b) cortical mechanisms specific for these tasks engage both hemispheres, and (c) in object construction only, men engage predominantly the right hemisphere whereas women show a left-hemisphere preponderance.

Modeling motor cortical operations by an attractor network of stochastic neurons

Understanding the neural computations performed by the motor cortex requires biologically plausible models that account for cell discharge patterns revealed by neurophysiological recordings. In the present study the motor cortical activity underlying movement generation is modeled as the dynamic evolution of a large fully recurrent network of stochastic spiking neurons with noise superimposed on the synaptic transmission. We show that neural representations of the learned movement trajectories can be stored in the connectivity matrix in such a way that, when activated, a particular trajectory evolves in time as a dynamic attractor of the system while individual neurons fire irregularly with large variability in their interspike intervals. Moreover, the encoding of trajectories as attractors ensures high stability of the ensemble dynamics in the presence of synaptic noise. In agreement with neurophysiological findings, the suggested model can provide a wide repertoire of specific motor behaviors, whereas the number of specialized cells and specific connections may be negligibly small if compared with the whole population engaged in trajectory retrieving. To examine the applicability of the model we study quantitatively the relationship between local geometrical and kinematic characteristics of the trajectories generated by the network. The relationship obtained as a result of simulations is close to the ‘2/3 power law’ established by psychophysical and neurophysiological studies.

A phenomenological theory of spatially structured local synaptic connectivity.

The structure of local synaptic circuits is the key to understanding cortical function and how neuronal functional modules such as cortical columns are formed. The central problem in deciphering cortical microcircuits is the quantification of synaptic connectivity between neuron pairs. I present a theoretical model that accounts for the axon and dendrite morphologies of pre- and postsynaptic cells and provides the average number of synaptic contacts formed between them as a function of their relative locations in three-dimensional space. An important aspect of the current approach is the representation of a complex structure of an axonal/dendritic arbor as a superposition of basic structures—synaptic clouds. Each cloud has three structural parameters that can be directly estimated from two-dimensional drawings of the underlying arbor. Using empirical data available in literature, I applied this theory to three morphologically different types of cell pairs. I found that, within a wide range of cell separations, the theory is in very good agreement with empirical data on (i) axonal–dendritic contacts of pyramidal cells and (ii) somatic synapses formed by the axons of inhibitory interneurons. Since for many types of neurons plane arborization drawings are available from literature, this theory can provide a practical means for quantitatively deriving local synaptic circuits based on the actual observed densities of specific types of neurons and their morphologies. It can also have significant implications for computational models of cortical networks by making it possible to wire up simulated neural networks in a realistic fashion.

A simulated actuator driven by motor cortical signals

One problem in motor control concerns the mechanism whereby the central nervous system translates the motor cortical command encoded in cell activity into a coordinated contraction of limb muscles to generate a desired motor output. This problem is closely related to the design of adaptive systems that transform neuronal signals chronically recorded from the motor cortex into the physiologically appropriate motor output of multijoint prosthetic limbs. In this study we demonstrated how this transformation can be carried out by an artificial neural network using as command signals the actual impulse activity obtained from recordings in the motor cortex of monkeys during the performance of a task that required the exertion of force in different directions. The network receives experimentally measured brain signals and recodes them into motor actions of a simulated actuator that mimics the primate arm. The actuator responds to the motor cortical commands with surprising fidelity, generating forces in close quantitative agreement with those exerted by trained monkeys, in both the temporal and spatial domains. Moreover, we show that the time-varying motor output may be controlled by the impulse activity of as few as 15 motor cortical cells. These results outline a potentially implementable computation scheme that utilizes raw neuronal signals to drive artificial mechanical systems.

A neural network learns trajectory of motion from the least action principle

This paper considers the problem of training layered neural networks to generate sequences of states. Aiming at application for situations when an integral characteristic of the process is known rather than the specific sequence of states we put forward a method in which underlying general principle is used as a foundation for the learning procedure. To illustrate the ability of a network to learn a task and to generalize algorithm we consider an example where a network generates sequences of states referred to as trajectories of motion of a particle under an external field. Training is grounded on the employment of the least action principle. In the course of training at restricted sections of the path the network elaborates a recurrent rule for the trajectory generation. The rule proves to be equivalent to the correct equation of motion for the whole trajectory.

A network analysis of developing brain cultures

We recorded electrical activity from four developing embryonic brain cultures (4-40 days in vitro) using multielectrode arrays (MEAs) with 60 embedded electrodes. Data were filtered for local field potentials (LFPs) and downsampled to 1 ms to yield a matrix of time series consisting of 60 electrode × 60 000 time samples per electrode per day per MEA. Each electrode time series was rendered stationary and nonautocorrelated by applying an ARIMA (25, 1, 1) model and taking the residuals (i.e. innovations). Two kinds of analyses were then performed. First, a pairwise crosscorrelation (CC) analysis (±25 1 ms lags) revealed systematic changes in CC with lag, day in vitro (DIV), and inter-electrode distance. Specifically, (i) positive CCs were 1.76× more prevalent and 1.44× stronger (absolute value) than negative ones, and (ii) the strength of CC increased with DIV and decreased with lag and inter-electrode distance. Second, a network equilibrium analysis was based on the instantaneous (1 ms resolution) logratio of the number of electrodes that were above or below their mean, called simultaneous departure from equilibrium, SDE. This measure possesses a major computational advantage over the pairwise crosscorrelation approach because it is very simple and fast to calculate, an important factor for the analysis of large networks. The results obtained with SDE covaried highly with CC over DIV, which further validates the usefulness of this measure as a computationally effective tool for large scale network analysis.

Blocking NMDAR Disrupts Spike Timing and Decouples Monkey Prefrontal Circuits: Implications for Activity-Dependent Disconnection in Schizophrenia

We employed multi-electrode array recording to evaluate the influence of NMDA receptors (NMDAR) on spike-timing dynamics in prefrontal networks of monkeys as they performed a cognitive control task measuring specific deficits in schizophrenia. Systemic, periodic administration of an NMDAR antagonist (phencyclidine) reduced the prevalence and strength of synchronous (0-lag) spike correlation in simultaneously recorded neuron pairs. We employed transfer entropy analysis to measure effective connectivity between prefrontal neurons at lags consistent with monosynaptic interactions and found that effective connectivity was persistently reduced following exposure to the NMDAR antagonist. These results suggest that a disruption of spike timing and effective connectivity might be interrelated factors in pathogenesis, supporting an activity-dependent disconnection theory of schizophrenia. In this theory, disruption of NMDAR synaptic function leads to dysregulated timing of action potentials in prefrontal networks, accelerating synaptic disconnection through a spike-timing-dependent mechanism.