**3. Implications for cognitive control**

## **3.1 Hippocampal properties are consistent with a role in cognitive control during volitional movements**

Cognitive control has variously been defined as a psychological construct for the coordination of thoughts and actions under conditions of uncertainty [98], the collective processes that organize different thoughts and memories, allowing the separation of currently relevant and irrelevant information [40], brain processes involved in regulating behavior according to internal goals or plans [38], and the ability to coordinate multiple streams of information to prevent confusion and select appropriate behavioral responses, especially when presented with competing alternatives [39]. Cognitive control processes allow us to efficiently process information and generate appropriate responses.

The essence of cognitive control is that neural processes involved in cognitive (psychological) processes act upon those regions of the brain needed to translate our thoughts into action. In this sense, "action" includes processes involved in making decisions, accessing memories, attentional control, response inhibition, and mental computations – i.e., any process that can potentially result in changes in behavior due to ongoing mental activity. Early explorations used tasks that unambiguously require cognition, such as:


Neural activity that correlates with these task behaviors is observed in prefrontal cortex [99–102], and prefrontal lesions in animals [103, 104] or humans [105] impair task performance. Neurological disorders associated with impaired frontal function, such as schizophrenia, also show impairments on these tasks [106]. The prefrontal cortex exerts top-down influences on sensory areas by functioning as a filtering mechanism that biases bottom-up sensory information toward the optimal response for a given context [107, 108]. This feedback loop may also be involved in memory recall, since the act of remembering evokes activity in the same sensory areas as the original event [96, 97]. The accumulated evidence supports a role of the prefrontal cortex in cognitive control.

Nonetheless, not all aspects of complex cognition benefit from prefrontal regulation, and the prefrontal cortex is not the sole source of cognitive control [109–111]. Indeed, the role for the prefrontal cortex in some functions may be limited. Prefrontal influence on activity in the primary motor cortex is indirect via dorsal

premotor cortex when learning to perform sequential finger movements, and absent during repetitive movements [112–114]. However, repetitive finger movements paced by a metronome do reflect cognitive control, as movements soon anticipate the auditory cue [49]. Under these conditions, the hippocampus provides connectivity selectively to the sensorimotor cortex hand representation, as shown in **Figure 1**.

An unequivocal set of criteria for a role in cognitive control has never been established, most studies relying on correlations between cognitive task performance and neural activity. In proposing minimal criteria for the cognitive control of movements [38], Burman noted an analogy between the skeletal movement system and the frontal eye field (FEF), which plays a critical role in volitional eye movements [115, 116]. Modifiable by cognitive influences, FEF cells only have three response properties, providing the basis for his proposed criteria:


Hippocampal connectivity with sensorimotor cortex during the repetitive tapping task arguably meets these criteria. Topographical connectivity maps were identified from single-voxel functional seeds that differentiated between movements of adjacent fingers [38]. Using the finger representations identified from an earlier study [117], the intensity of connectivity from each functional seed was then compared statistically across time periods for movement of each finger. The criteria for cognitive control were met:


The extent to which these results can be generalized is limited, as the spatial area covered by finger movements was restricted, only one temporal pattern of movements was tested, and only finger movements were involved. Nonetheless, known hippocampal properties are consistent with these conclusions: hippocampal function is associated with conscious states such as declarative memories, and the hippocampus is sensitive to the timing of events as well as their spatial properties. Exploring a wider range of spatial regions involved in physical manipulation, as well as varied durations of movements, could more fully delineate the extent of its cognitive control over volitional movements.

*Hippocampal Influences on Movements, Sensory, and Language Processing: A Role in Cognitive… DOI: http://dx.doi.org/10.5772/intechopen.100122*

### **3.2 Evidence for cognitive control suggested by prefrontal and hippocampal connectivity studies**

**Figure 6** summarizes cortical connections between areas involved in sensory, language, memory, and motor functions. Sensory input from primary sensory cortices passes to association cortex for feature extraction (yellow arrows), then to higher centers involved in language, memory formation and cognition. The higher centers have bidirectional connections (blue arrows); in addition, the hippocampus and prefrontal cortex modify activity in sensory, language, and motor areas (red arrows). As described below, the red arrows represent candidates for cognitive control.

Cognitive control requires influence from a higher center to modify neural activity in those areas required to perform a task. The prefrontal cortex and hippocampus [40] have both been suggested to play such a role, and both have connections appropriate for a role in cognitive control over sensory input, language, and memory (as shown in **Figure 6**). Connectivity from prefrontal cortex to motor areas is indirect and limited to sequential movement learning tasks; otherwise, there is little in the pattern of connectivity to differentiate between these candidates. The hippocampus and prefrontal cortex show strong interactions, suggesting they may often work jointly to exert cognitive control. Such redundancy would have an evolutionary advantage, since damage to either area by itself will be less crippling. As shown in **Figure 6**, hippocampal effects on sensory areas may also be indirectly mediated through prefrontal cortex. This could explain a number of curious findings, such as why the hippocampus is needed for memory formation (to provide contextual information to prefrontal cortex) but not for memory recall (when prefrontal cortex provides memory recall by reactivating sensory areas involved in sensory

### **Figure 6.**

*Summary diagram of cortical connections. Sensory pathways are colored yellow, modulatory feedback pathways red, and bidirectional connections blue. Red pathways in this summary originate from prefrontal cortex and the hippocampus, which are both likely to play a role in cognitive control, informed by their mutual connections and feedback.*

perception). Such redundancy can also account for residual cognitive abilities during neurological disorders that disrupt function in either region.

A recent study compared functional connectivity during tasks and the resting state condition, the latter representing the intrinsic architecture of the brain [119]. Small but consistent changes were observed across dozens of task states, suggesting both task-specific and task-general network changes. Appearing within the resting state network, the hippocampus (but not prefrontal cortex) accounted for most variance in connectivity across all tasks. This finding suggests that the hippocampus, unique from prefrontal cortex, plays a primary role in regulating task behavior.

### **3.3 Testing for cognitive control in the hippocampus and prefrontal cortex**

Regardless of the brain region studied, the same criteria can be applied to establish a role in cognitive control. It was previously noted that early studies used tasks that unambiguously required cognition to perform, such as the Stroop and n-back tasks; prefrontal cortex was implicated because response properties correlated with task performance. This approach addresses the first of the three criteria proposed above: the neural activity is tied to a cognitive or volitional state of consciousness. For the hippocampus, the need to relate hippocampal activity to a cognitive or volitional state may not initially be apparent due to its association with declarative memory, yet hippocampal activity has also been reported during the formation of implicit memories [120–122]. Until we know the functional role of hippocampal activity during implicit learning conditions (decision-making? context? association pairings?), we cannot assume that hippocampal activity necessarily reflects a volitional state of consciousness. Behaviors acquired through hippocampus from repeated stimulus–response associations, for example, require little thought and may ultimately be mediated by the cerebellum and striatum [48, 123–125].

Involvement in cognitive processes does not in itself indicate a role in cognitive control. The left inferior frontal gyrus (Broca's area) was activated by words in the Stroop task, for example, yet the behavioral response suggested by the meaning of conflicting words had to be suppressed for accurate performance. To play a role in cognitive control, brain regions involved in cognition must act upon brain areas required to achieve the goal of a task. Effective connectivity tools provide a method to study such effects, particularly useful when demonstrable effects are taskspecific. (This is the advantage of the PPI technique, which is both directional and task-specific).

Such a task-specific influence must be relevant to task performance. This is the purpose of the temporal- and spatial-selectivity criteria suggested above: task performance is always delimited in time, and typically involve perceptual stimuli or volitional movements appearing within a spatial environment. Cognitive processes such as emotional associations may not be invariably linked to a concrete stimulus, yet the provocative stimulus in an experiment can still be spatially delimited. Any additional constraints imposed by a task should also be reflected in a signal for cognitive control.

A role in cognitive control can be confirmed when loss of the control signal results in the inability to perform the task. As noted above, however, the behavioral deficit will not be complete unless signals are removed from all areas involved, which may include both the hippocampus and prefrontal cortex.

### **4. Conclusions**

The hippocampus provides task-specific influences on sensory, motor, language, and mnemonic areas of the brain. Detailed analysis in sensorimotor regions during

*Hippocampal Influences on Movements, Sensory, and Language Processing: A Role in Cognitive… DOI: http://dx.doi.org/10.5772/intechopen.100122*

motor tasks showed a pattern of connectivity consistent with the requirements for cognitive control; connectivity patterns across tasks suggest a joint role for the hippocampus and prefrontal cortex. Criteria and suggestions for further evaluation of cognitive control from both regions are considered.
