Scientists can record the activity of single neurons in the monkey brain. One type of neuron, called a mirror neuron (MN), fires when the monkey performs a certain action and also when it observes another person doing that same action.
MNs may serve to enable the monkey to understand actions and goals and copy others’ movements. For example, MNs in the premotor cortex and supplementary motor areas have been shown to discharge when monkeys see their own hands reaching for objects like tool pliers.
Fast cortical oscillations in the beta band (12-25 Hz) occur spontaneously in the primary motor (M1) and supplementary motor areas (SMAs) of awake monkeys. Oscillations typically occur in episodes of 2-5 oscillatory cycles. These oscillations can be triggered by a range of visual and auditory stimuli. The responses of multiple spatially distributed neurons to natural scenes are ordered in predictable sequences. The rank order of these response sequences is maintained when the stimulus parameters are varied. This suggests that natural stimuli trigger a matching operation between sensory evidence and priors stored in the cortex.
To study the role of these oscillations in neocortical short-term Hebbian plasticity, ECoG recordings were collected from multiple sites in sensorimotor and SMA cortex in two awake monkeys using epidural electrodes. Stimuli evoked cortically evoked potentials at each site, and the average of 30
The amplitude of the CEP triggered by a specific stimulus at each M1 or SMA site was found to be correlated with the instantaneous phase of the beta cycle at that site when the stimulation was delivered. When the stimulation was delivered during a phase of the beta cycle that corresponded with the peak at a particular site, there was a large conditioning effect on the evoked volley at that site. This was characterized by potentiation during depolarizing phases and depression during hyperpolarizing phases.
Similar conditioning effects were observed at non-triggering sites that exhibited oscillations synchronized with those at the triggering site. However, the magnitude of these changes was less than those at the triggering site and was inversely related to the phase difference between the triggering and non-triggering sites.
These findings suggest that beta oscillations in neocortical pyramidal cells are the substrate for Hebbian plasticity. Specifically, they provide a temporal and spatial template for synchronization between pre- and postsynaptic activity and a dynamically adaptive memory system in the brain. It is thought that these mechanisms are important in the function of mirror neurons in human and monkey premotor cortex that respond to hand actions and other objects, and in the Broca’s area of human speech and language regions that are homologous to ventral premotor cortex.
When stimulation of premotor cortex or the supplementary motor area is applied for longer periods (500 msec), it is able to elicit more complex movements than stimulation of primary motor cortex. This indicates that these higher level areas have the ability to encode more complicated patterns of movement than can be evoked by primary motor cortex alone, and to select appropriate motor plans to achieve desired end results.
For example, when a monkey is trained to move an arm to different locations that vary in both direction and distance from the target, recordings show that many primary motor cortex neurons exhibit a characteristic bell-shaped speed profile. They accelerate during the initial portion of the movement, reach a peak velocity approximately half way to the target, and then decelerate until they reach their destination (Figure 3.9).
In another study, a monkey was trained to make a certain movement in response to a visual signal with a variable delay between the onset of the signal and the initiation of the movement. Recordings of premotor cortex neurons show that these neurons fire selectively during this delay interval. They fire to reflect the monkey’s perception of a person performing the same action as the one that the monkey is about to make.
Other studies have shown that the sensory thalamus has important roles in perceptual performance and decision making. For instance, when the monkey is presented with a visual stimulus that must be detected versus a tactile stimuli, the responses of sensory thalamus neurons differ significantly. This demonstrates that the sensory thalamus has an integrated representation of the stimulus characteristics that is used to detect its presence or absence, regardless of whether it is perceived as a visual or tactile stimulus.
The amplitude modulation of VPL neuron firing rate also reflects the efficiency of detection. To test this, we conducted a series of experiments in which the peristimulus time histograms of VPL neurons were measured and fitted to a logistic curve to determine the stimulus amplitude at which 50% of hits and CRs were observed. The resulting psychometric and neurometric curves were very similar, but the amplitude modulation of the neurometric curve was smaller than that of the mean psychometric curve. This indicates that VPL neurons use a simple and robust neural code for stimulus amplitude, rather than relying on the periodicity of spike train coding.
Neurons that respond to a particular category of visual stimuli can be classified according to their response sequence, rate vectors, and polarity. This classification is used in decoding studies to extract perceptual information from neuron responses. Decoding experiments have shown that neurons that are able to distinguish between natural and manipulated stimuli display higher selectivity, implying that these neurons can extract behaviorally relevant information from sensory evidence.
In the basic observation task, monkeys were instructed to keep their hand (contralateral to the recorded hemisphere) in a resting position and observe different videos displayed on a video monitor. These videos included a monkey gripping an object seen from a first or third person perspective, a human grasping an object seen from a lateral view or mimicking this action, a biological movement (BM), and the motion of an object (OM).
Each trial consisted of a series of 200 ms videos with an interstimulus interval of 500 ms. The stimulus intensity increased in a ramp-like fashion starting from zero and reached its maximum value within 2200 ms, after which it remained at the same level for 500 ms (plateau time). The durations of the pre-stimulus baseline and the plateau phase were randomized to reduce trial-to-trial variability.
The results of single-neuron activity during the observation of the different videos showed that all neurons discharged at least during a portion of one of the epochs of the videos presented in each condition. Out of these, 21 neurons responded to videos depicting a monkey gripping an object from a first or third person perspective and 14 responded to videos depicting a human grasping or mimicking this action.
In contrast, only 12 of these neurons discharged also during the observation of the BM or OM videos. Interestingly, the responses of these neurons to the HG and HM videos were the strongest during the first epoch, which was the earliest that the monkey was gripping an object in the respective video.
The fact that many VLPF neurons are selective for biological stimuli, and exhibit a preference for the first epoch of videos depicting the action performed by the monkey, suggests that these neurons code the earliest available information for predicting whether this is a functionally relevant gesture. For this reason, the coding of these behaviors in prefrontal cortex has been termed ‘simulus-specific’ and a ‘simulus-invariant’ orienting mechanism.
The premotor cortex sends axons to both the primary motor cortex and directly to the spinal cord, and performs more complex task-related processing than the primary motor cortex. Stimulation of the premotor cortex with a high level of current often results in movements that are not directly related to the initial starting position of a body part, and may instead require a change of posture or positioning (Figure 3.2). These neurons are called “motor-set” neurons because they appear to play a role in selecting the appropriate movement to achieve the desired end result of the task.
Studies examining monkey neuron activation have shown that some VLPF neurons are sensitive to the observation of biological and non-biological movements, but that their responses to the observations are not necessarily related to the selection of a particular movement or even the general type of movement being observed. For example, a neuron described in Figure 2a responds strongest to the observation of a monkey grasping an object from either the first or third person perspective. Its discharge starts before movement onset, peaks during the grasping action and continues until the object is touched. However, this neuron does not show a significant increase in its discharge to the observation of a second video that shows a second grasping action from either the first or third person perspective, and it actually decreases its response to the latter video.
Another interesting finding is that some primary motor cortex neurons encode the speed of a targetted movement. Almost all targeted movements follow a typical bell-shaped curve, where the hand accelerates to reach a certain point, then decelerates to a steady velocity that will bring it to the desired location. The firing rate of many neurons in the primary motor cortex correlates with this movement speed, suggesting that these neurons are able to record this information.
In a study employing optogenetics, researchers have been able to activate monkey neuron activation by delivering a low-level light stimulus to the somatotopically identified forelimb region of primary motor cortex using a homemade optrode with channelrhodopsin-2 (ChR2) expressed under control of a strong ubiquitous CAG promoter. The injected neurons were then stimulated with short-duration optogenetic intracortical microstimulation to record their responses to videos of different biological and non-biological movements.