What is the relationship between visual attention network and visual stream (Two-streams hypothesis)?

Question

There are two visual attention networks proposed by Maurizio Corbetta and Gordon L. Shulman (2002). These are top-down and bottom-up attention networks.

One system, which includes parts of the intraparietal cortex and superior frontal cortex, is involved in preparing and applying goal-directed (top-down) selection for stimuli and responses. This system is also modulated by the detection of stimuli. The other system, which includes the temporoparietal cortex and inferior frontal cortex, and is largely lateralized to the right hemisphere, is not involved in top-down selection. Instead, this system is specialized for the detection of behaviourally relevant stimuli, particularly when they are salient or unexpected.

There are also two ventral streams proposed by Melvyn A. Goodale and A. David Milner (1992). These are ventral stream and dorsal stream.

We propose that the ventral stream of projections from the striate cortex to the inferotemporal cortex plays the major role in the perceptual identification of objects, while the dorsal stream projecting from the striate cortex to the posterior parietal region mediates the required sensorimotor transformations for visually guided actions directed at such objects.

My question: What is the relationship between these two proposed networks? As in do they affect each other, or is one a preliminary network that occurs before the other? There have been studies on this but I am having difficulty making sense of them.

Example: Visual streams and shifting attention (Brown 2009) http://www.sciencedirect.com/science/article/pii/S0079612309176045

To gain insight into the relationship between M/dorsal and P/ventral activity and deploying visual attention, we used a covert cuing paradigm to manipulate attention while bottom-up and top-down perceptual stimulus variables created M/dorsal and P/ventral-biased conditions. One study examined the object advantage, where responses are faster for within-relative to equidistant between-object shifts of attention. Visual stream contributions to object- and spaced-based attention were revealed using psychophysically equiluminant conditions expected to reduce M/dorsal activity. Other studies investigating visual stream contributions to location-based inhibition of return (IOR) used IOR magnitude as an indicator of the ease or difficulty of deploying spatial attention. Greater IOR was found under P/ventral-biased conditions. Less IOR was found under M/dorsal-biased conditions.

Answer 1

My short answer would be that there is no clear relationship between these two models about how visual information processing works in the brain. However, the terminology and evidence overlaps considerably which makes this confusing.

One way to think about this is that the two models are explaining different stages of processing. The two visual streams is somewhat more general and it is very much founded on brain anatomy (lesion studies in animals and neuropsychology). It is to do with what happens to the visual information that comes into the brain, and the theory is that while some of it is processed to consciously recognise stuff (ventral), a separate pathway is involved with supplying visual information for action.

None of this is explicitly about "Attention". Corbetta and others are trying to relate brain structures to experiments in typical observers where processing changes even though stimuli are perceptually available. Although their two networks are sometimes called "dorsal and ventral", because of the involvement of superior/inferior brain regions, this does not imply that they map on to Milner and Goodale's visual streams. There are several other differences in anatomy.

It seems likely that both top-down and bottom-up attention can modulate perceptual processing in both visual streams (and they interact too!). It might help to think of some examples...

• If you are more likely to act on something when it is more salient (e.g., avoiding a colourful obstacle without thinking about it) then this suggests the "bottom-up" system has detected the obstacle and prioritised it's processing for action (dorsal stream).

• If you are voluntarily told to pay attention to something, and you are quicker to act on this, it suggests that "top down" attention has enhanced the dorsal stream.

• If you are voluntarily told to pay attention to something and you are quicker to recognise it, then this would be an effect on conscious
  recognition, thought to be part of the ventral stream.

• If your attention is captured "bottom-up" by something, this is also going to make you better at consciously recognising it with your
  ventral stream.

Here, and here are a couple of recent neuroscience papers on the attentional systems involved, and here is a classic paper showing attention is coupled in both perception and action streams.