Visual search is an important cognitive ability that allows us to respond to the world around us. Current theories suggest visual search may be controlled in part by the dopamine system. As patients with Parkinson's Disease have dysregulation of the dopamine system they can provide useful insights into how this system works.
Visual Search Deficits in Parkinson's Disease
Think of your eye as a digital camera. A typical digital camera might have a resolution of say 5 megapixels, that is, 5 million tiny pixels, each of which contains a piece of information about a block of space. As you may have experienced, the higher the resolution of the picture you take, the more space it takes up on your memory card or hard disk. An image shot at 3 megapixels might take up 2.5mb whilst an image at 5 megapixels might take up 5mb for instance. The resolution of the human eye is around 370 megapixels, and rather than taking individual pictures, it is more like a high-definition video camera. The "storage space" it would take to permanently store all this information would rapidly become unmanageable. Your entire brain capacity would rapidly be filled up with irrelevant information like how many leaves there were on a tree or every piece of text that you've ever read. So one of the most important things a brain must be able to do is discard huge amounts of information all the time in order to only process and react to what is relevant. To cast things back to thousands of years ago when humans lived in the African savannah, if you're walking through the desert and you come upon a snake, it would be useful for some part of your brain to make the snake in front of you a higher priority than say, the number of rocks you can see in the background.
Visual search in the brain
Horowitz et al (2006) stated: "Visual search refers to a wide range of tasks where an individual must search for a target item (whose location is uncertain) in the presence of one or more distractor items." A real world example might be searching for a jigsaw piece or for your wife in a crowded shop. Whilst we can see many items in our field of vision, we can generally only concentrate on one item at a time. This is a good description of what we call "attention", which acts as a kind of spotlight for things we are looking at. To a certain extent, our brain automatically assigns priority to different objects. Those that are thought to be of high priority are said to be "salient" or have high "salience". The brain is thought to use a salience "map" to assign higher priority to some objects than others in order to minimise the number of items that need to be searched for. This can be done in two ways. "Bottom-up" information consists of the difference between the item you are searching for and any distractors along a particular set of dimensions, so if your wife is very tall and she is surrounded by very short women she should be easier to spot. "Top-down" information consists of what you know about the target already, so, if you're looking for your wife who has a red coat on then you know to exclude all women who are not wearing a red coat from your search. Both sorts of information each weigh the salience map of what you see in front of you to make some things "stand out" from other items (the distractors).
Links with Parkinson's
What's this got to do with Parkinson's Disease? We know that PD causes degeneration of the substantia nigra, the part of the brain that produces dopamine. In some parts of the brain, dopamine is involved in initiating movement. However it has many other roles too. Dopamine has been implicated as being important in cognitive abilities such as switching between different rule structures or concentrating attention to a subject. Dopamine may play a role in selecting which types of information received by the brain are important. It may serve to amplify those signals which are relevant, whilst diminishing those that are not. Therefore dopamine may act as a sort of "information filter" in determining which sources of input in the environment merit attention. In this context then, Parkinson's Disease is being used as a sort of "natural experiment" to see what it looks like when you compare people with a normally functioning dopamine system and a dysfunctional dopamine system. Although we do learn something about Parkinson's Disease in itself, the main point is to look at the function of dopamine in the human brain more generally.
In a recent study, Horowitz et al used a number of experiments to investigate this phenomena in Parkinson's Disease. Given what we have said about dopamine there are a number of predictions that could be made. A depletion of dopamine should cause a less effecient visual search, as there is no longer a system amplifying salient signals and reducing irrelevant signals. Therefore patients with Parkinson's Disease might find it more difficult to find a target item amongst a set of distractors. Horowitz et al were interested in whether the difficulties experienced were due to problems with "bottom-up" information or "top-down" information. After all, patients with Parkinson's Disease often benefit from external cues, such as walking on alternately coloured tiled surfaces or walking over lines on the ground. Previous research studies have varied somewhat in their findings, and haven't always had a clear, testable idea in mind about how visual search should work in healthy participants. Horowitz et al state that in summary, previous research appeared to show that "a.) there is little evidence for a deficit in the speed of shifting attention between items" and b.) (patients with Parkinson's Disease) are more likely to demonstrate impaired performance when there is uncertainty about the identity of the target and selection must be guided entirely by "bottom-up" factors.
Horowitz et al's (2006) experiments
For their study, Horowitz et al compared 10 moderate-to-severe Parkinson's Disease patients with 10 matched (i.e. very similar) healthy participants on visual search tasks in which the level of either "bottom-up" or "top-down" salience could be manipulated. Although the Parkinson's Disease patients were on medication at the time, which consists of dopamine-replacement therapy, the authors suggest that medication does not completely replace absent dopamine in the brain. The task consisted of a computer screen where participants had to say whether or not they could see a particular target amongst a set of distractors. The stimuli were always either horizontal or vertical bars, which would always be either red or green. There were three levels of complexity for "bottom-up" information, illustrated in the diagram below. In the simplest task, "feature search", all the distractors were the same (e.g. red horizontal bars) and the target was a red vertical bar. In the "noisy feature search", the distractor lines could be either red or green, but the target remained a red vertical bar. Finally, in the "conjunction search", the distractors each had one feature in common with the target (i.e. red horizontal bars or green vertical bars). There were two levels of manipulation for "top-down" information i.e. what the participant was told about the target. In "Target known" they were told that the target would always be a red vertical bar for every trial of the test. In the "Target unknown" condition they were simply told to identify the odd one out. The main outcome measure was the reaction time that it took each participant to accurately identify whether or not the target was there. In half of the trials there was a target, in half there wasn't.
In terms of "top-down" processing, in the "Target known" conditions, Parkinson's Disease patients performed normally, regardless of the complexity of the "bottom-up" information provided. However in the "Target unknown" conditions, Parkinson's Disease patients were significantly slower across the board, with increasing deficits as the complexity of "bottom-up" information increased; in the most complex "conjunction search" task, Parkinson's Disease patients took twice as long as healthy participants to correctly identify the target. So, there was no deficit in terms of identifying which objects were salient, nor was there evidence for impaired ability to shift attention between objects. The authors suggest that the problem lies in what patients do with the results of the search that matters. Cognitively, the problem may lie either in visual short term memory (i.e. when searching for a target that was different to the others, they had difficulty remembering what all the others had looked like), or in terms of decision and response processing (i.e. once the target had been identified, what they had to do next). The results also indicate that giving patients with Parkinson's Disease as much "top-down" information as possible can allow them to function as effeciently as healthy people without the disease, and this provides experimental evidence that external cues which limit the need to make decisions or choose from a variety of responses may help facilitate actions.
The authors have established that in terms of how people with Parkinson's Disease see the world, there is no deficit with the way their brain processes some items as being more important (or salient) than others. In a simple setup with a known target, the desired item does "pop out" as it does for healthy people. However, when patients are not explicitly told what they should be looking for, they exhibit significant deficits. Therefore we can say something about the role of dopamine in the visual search behaviour of the healthy brain. The authors hold that it is not involved in the initial selection of salient objects. However, they suggest that an absence of dopamine "seems to interfere with the conversion of a saliant visual signal into action". They suggest that future research might look at how "top down" information can be maximised and "bottom up" information minimised as a tool in therapy or rehabilitation.
Click to read this topic
6/6/2012 2:43:17 PM - 184.108.40.206