The thalamus is the switchboard of the brain. The reticular formation regulates excitability and forms a sheet around the thalamus and can gate all information (high arousal) or block all information (coma). Arousal is a prerequisite for attention. Selective attention is a covert, cognitive brain mechanism which enables us to process relevant information, thoughts and actions and ignore or suppress irrelevant or distracting input, thoughts and actions. Selective attention can be voluntary or involuntary. Overt attention is attention located at the same location as the eyes. Covert attention is attention located at another location than the eyes.

Voluntary attention (top-down) is regulated by the dorsal frontoparietal network. It includes the superior parietal lobe (SPL), frontal eye fields (FEF) and the middle frontal gyrus (MFG). Involuntary attention (bottom-up) is regulated by the ventral frontoparietal network. It includes the temporoparietal junction (TPJ), inferior frontal gyrus (IFG) and the middle frontal gyrus (MFG). This network is strongly lateralized to the right hemisphere.

Subcortical structures are important for different aspects of attention. The superior colliculus is involved in moving attention. The temporoparietal junction is involved in disengaging attention and the pulvinar is involved in engaging attention.

Neglect involves the inability to disengage attention and occurs because of right-hemisphere lesions to the temporoparietal junction or the inferior frontal gyrus (involuntary attention) and is often temporary. Extinction, focussing on the contralateral side of the lesion when both visual fields are stimulated often remains.

Balint’s syndrome involves simultagnosia, only seeing one object at a time, optic ataxia, trouble with visual guidance for reaching objects and oculomotor apraxia, problems in making voluntary eye movements to objects and occurs because of bilateral lesions to the dorsal posterior parietal lobe and the lateral occipital complex (LOC).

The capacity of selective attention is 3-4 items and selective attention is relatively slow. It leads to processing at the expense of something else. This limited capacity is caused by the voluntary attention network. The attentional blink refers to missing a second target if it is presented about ±300ms after the first target because attention is still processing the first target.

In the Posner cueing task, participants have to react to a cue. The central cue triggers voluntary attention and the peripheral cue triggers involuntary attention.

A neutral cue is used to measure the baseline reaction time. Reaction time increases with a valid cue and reaction time decreases with an invalid cue and the time between the cue and the target needs to be at least ±200ms in the central cue condition. The cue validity is important. If the cue validity is too low, people stop using the cue.

In the peripheral cue condition, attention is reflexive and fast. If the cue-target interval is below ±200ms, there is facilitation of attention. If the cue-target interval is above ±200ms, there is inhibition of attention.

The feature integration theory of attention states that elementary stimulus features can be processed pre-attentively and in parallel and that spatial attention must be directed to relevant stimuli in order to integrate the features into the perceived object and this must be deployed sequentially. This theory explains why people are quicker in the pop-out condition than in the conjunction search condition.

The more salient a stimulus is, the more easily our attention is captured. Inhibition of return (IOR) refers to responding more slowly to a location where attention had just been. It can be overridden by using voluntary attention.

There is a positive ERP (P1) around ±100ms after stimulus presentation. P1 is more positive when spatial attention is at the correct location when the stimulus appears. It reflects the inhibition of return effect and there is a P1 topography. There is a negative ERP (N2PC) around ±200ms after the stimulus presentation. N2PC is apparent when spatial attention moves from the right to the left (or v.v.). It is the strongest contralateral to the attentional movement.

The brain uses the largest receptive field possible. The receptive field of attention decreases with the processing stage. The earlier something is processed, the smaller the receptive field. Receptive fields become larger when going up in the visual hierarchy. A receptive field can only represent one object at a time. Multiple objects can fall within the receptive field of a neuron and this produces competition between objects for representation. Attentional effects are strongest in V4 and become weaker in the earlier stages of processing. The neural site of spatial attention depends on the specificity of the target (required receptive field).

An effective stimulus drives the action potential of a neuron maximally. The neuron acts as if it only sees the effective stimulus. There is no competition or effect of attention in V1. There is competition in V4 if objects are presented simultaneously. This leads to more competition, which leads to less activity. The biased competition model states that competition in V4 is resolved by directing attention to a specific stimulus.

Feature-based attention includes attention to features and enhances activity in the brain region designated to process the attended feature. There is object-based attention. There is inhibition of non-attended stimuli. Certain neurons have a preference for a specific feature. Simple cells exhibit orientation tuning and respond to contrast borders.

The early selection model of attention states that there is early processing of some information and no processing of non-attended information. The late selection model of attention states that all information is processed and selection occurs afterwards. The attenuation model of attention states that either early or late selection depends on the amount of attention deployed. If a lot of attention is deployed early selection occurs. If little attention is deployed, late selection occurs. Early selection is used during very difficult tasks and when you know what to attend to. Late selection is used during easy tasks and when you don’t know what to attend to.