Same, Same but Different? A Multi-Method Review of the Processes Underlying Executive Control

Abstract

Attention, working memory, and executive control are commonly considered distinct cognitive functions with important reciprocal interactions. Yet, longstanding evidence from lesion studies has demonstrated both overlap and dissociation in their behavioural expression and anatomical underpinnings, suggesting that a lower dimensional framework could be employed to further identify processes supporting goal-directed behaviour. Here, we describe the anatomical and functional correspondence between attention, working memory, and executive control by providing an overview of cognitive models, as well as recent data from lesion studies, invasive and non-invasive multimodal neuroimaging and brain stimulation. We emphasize the benefits of considering converging evidence from multiple methodologies centred on the identification of brain mechanisms supporting goal-driven behaviour. We propose that expanding on this approach should enable the construction of a comprehensive anatomo-functional framework with testable new hypotheses, and aid clinical neuroscience to intervene on impairments of executive functions.

Glossary

‘Attentional control of action’ model

Norman and Shallice (1986) proposed the existence of two attentional control processes: (a) the contention scheduling mechanism that selects between conflicting routine actions, and (b) the supervisory attentional system (SAS) that intervenes when the selection of routine actions is not sufficient (as in novel situations). The authors suggested that a deficit of a supervisory system involved in the control of action generates a dysexecutive syndrome.

Attentional episodes

Effective cognition requires the sequential resolution of hierarchically organized subproblems, each one with its own relevant inputs, actions, and outcomes. A step-by-step implementation (i.e., episode) of selective attention supports cognition by admitting into the current task set only relevant information (Duncan, 2013).

Attention Networks Model

First proposed by Posner and Petersen (1990) and later revised in 2012 (Petersen and Posner), this model identifies three independent but interactive attentional components of alerting (defined as the ability to produce and maintain a state of readiness to respond to incoming stimuli), orienting (defined as the ability to select some aspects of the information to prioritize its processing), and executive control of attention (the ability to select the most appropriate response among competing responses and suppressing the effect distracting information)

Covert or overt attention

The allocation of attention towards a region of the visual field with (overt) or without (covert) eye movements.

Dorsal Attention Network (DAN)

A bilateral set of frontal and parietal regions showing concurrent activation increases during goal-driven attention (Corbetta & Shulman, 2002).

Decision making

The process of deliberating between two or more options that vary in terms of subjective value or environmental relevance.

Direct electrocortical stimulation (DES)

A neurosurgical method for functional brain mapping. A small electrical current applied directly to cortex or white matter, causing temporary interruption to the underlying network. If behaviour remains unchanged following stimulation, the structure being probed is considered to be non-essential and resectable (i.e., its resection is unlikely to result in permanent neurological deficit), but if there is a relevant behavioural change with stimulation, the structure is considered to be an essential network component and therefore non-resectable.

Default Mode Network (DMN)

A bilateral set of medial regions (including, but not limited to, the ventrolateral prefrontal cortex, and the posterior cingulate cortex) showing concurrent deactivation during task execution in a variety of fMRI or PET studies (Raichle et al., 2001).

Exogenous Attention

Often studied using non-predictive visual cues preceding an upcoming visual target, defines the allocation of attentional resources towards a salient stimulus through an automatic, effortless, process.

Endogenous Attention

Often studied using predictive visual cues preceding an upcoming visual target, defines the allocation of attentional resources towards a behaviourally relevant stimulus through a voluntary, effortful, process implemented in a top-down fashion.

Fluid and Crystallised intelligence

The two components of general intelligence relate to reflecting ability to solve problems by abstraction (fluid) and the ability to solve problems by using general knowledge and previous experience (Cattell, 1963).

Functional MRI

An MRI technique that indirectly measures activation of brain regions based on blood flow (BOLD activity) either during the completion of a task (task-based) or at rest (resting state).

Frequency bands

Initially discovered in EEG studies, refer to the rhythmic, repetitive activity of brain waves recorded by electrodes. More recently studied using also other electrophysiological and neurophysiological methodologies. This rhythmic activity is usually divided into bands, including delta (< 4 Hz), theta (4 to 8 Hz), alpha (8 to 15 Hz), beta (15 to 32 Hz), and gamma (> 32 Hz).

g factor

A general factor that summarizes an individual cognitive abilities, often measured using IQ tests such as the Wechsler Adult Intelligence Scale, first proposed by Spearman (1927).

Hub area

A brain region (or sub-component of a network) critically positioned within the network architecture of the brain due to their high number of contacts with other brain regions (Power et al., 2013).

iEEG

Intracranial electroencephalography, an invasive method for recording electrical activity from small populations of neurons in the cerebral cortex and in subcortical structures, performed in patients with intractable epilepsy to localise their epileptic focus. Includes two types of recordings: (a) Electrocorticography (ECOG), recording of cortical activity using electrodes placed directly on the brain surface; and (b) Stereotactic EEG, recording of cortical and subcortical activity using penetrating depth electrodes inserted into the brain (Mukamel & Fried, 2012; Parvizi & Kastner, 2018).

Inhibition

The active process of suppressing irrelevant stimuli, task sets, behavioural responses, and unwanted thoughts from capturing resource-limited attentional function and, consequently, affecting performance (Aron, 2007).

Lateralization

A principle of brain organization that allocates the implementation of one cognitive function exclusively (or for the major part) to one cerebral hemisphere (e.g., Heilman & Van Den Abell, 1979).

Mind wandering

A mental state whereby the focus of attention is internal, self -generated thought (see Mason et al., 2007 for details).

Multivoxel pattern analysis (MVPA)

A statistical technique that applies pattern classification algorithms to Blood Oxygenation Level Dependent (BOLD) activity acquired using functional MRI.

Neglect (hemispatial neglect)

Failure to spontaneously orient to one side of space. Most often presents as left neglect following right hemisphere damage (Heilman & Watson, 1977).

Neural oscillations

Rhythmic patterns of neural activity in the central nervous system that can occur during activation of small-to-large populations of neurons and appear to support cognitive processes.

Phasic alerting

Temporary increase in responsiveness produced by the presentation of a sudden stimulus acting as a warning signal (Aston-Jones & Cohen, 2005).

Task set

A stimulus–response relationship that includes formation of a criterion to respond to a target, organization of the program to complete a specific task, and adjustments of the automatic processes.

Tonic alerting

The ability to produce and maintain a prolonged state of readiness in the attempt to detect rarely occurring stimuli, often characterized by a monotonic decrease in performance as the required task duration increases. Closely related to the concepts of energization and sustained attention in other attentional models (Aston-Jones & Cohen, 2005).

VAN

A set of fronto-parieto-temporal regions lateralized to the right hemisphere showing concurrent activation increases during bottom-up or exogenous attention (Corbetta & Shulman, 2002).

Voxel-based lesion-symptom mapping (VLSM)

Univariate method of neuroimaging analysis comparing patients with or without lesions in any given voxel (or region of interest, for the volume of interest-based version of this method) with respect to differences in behavioural measures (t-statistic is computed per voxel: see Bates et al., 2003 for details).