Representation and processing of spatial information in lateral occipital cortex

Humans have a very sophisticated vision. Much of this sophistication lies within the brain where large expanses of cortex are devoted to vision. However, understanding how the human brain endows us with our visual abilities remains a significant challenge. Functional Magnetic Resonance Imaging (fMRI) has provided an informative way in which to gain traction on this issue by yielding measurements that allow the brain to be divided into different visual areas. Two broad approaches have been adopted: Firstly, regions within the visual cortex have been characterised in terms of how they respond to different kinds of stimuli. Research has revealed areas that are selectively responsive to properties such as colour and motion or object categories such as faces, places or everyday objects. Secondly, recording brain activity while participants view stimuli that move systematically through the visual field can determine which parts of the brain contain maps of the visual field. Such visual field maps are easy to detect in the early visual processing stages of the human brain and, as techniques have been refined, it has become clear that many more visual field maps can also be found at higher processing stages. Importantly, some areas that are selectively responsive to specific kinds of visual stimuli also contain multiple maps of the visual field. This selectivity is unlikely therefore to arise from the neural processing that is performed and replicated in each constituent visual field map. Instead, each map will perform one or more crucial roles in visual processing and contribute to the category selectivity found in the regions that encompass them. As yet, however, the account of what processing roles the multiple visual field maps play, and how they contribute to human visual perception, is far from complete.

The aim of this proposal is to study the lateral occipital area of the human brain, which is known to be specialised for visual object perception. This area of the brain has also been shown to contain six separate visual field maps (LO1-6). Our proposed work hinges on two very recent findings. Firstly, the six visual field maps that we wish to investigate have only relatively recently been discovered, but our lab is one of only a few that is able to reliably identify them. Secondly, rather than relying on fMRI alone to probe the function of visual areas of the brain, we have successfully used transcranial magnetic stimulation (TMS) to selectively disrupt processing in two of the visual field maps and have determined their causal roles in orientation and shape perception. Taken together, these findings give us the unique capability to evaluate how the visual field maps LO1-6 process and represent visual information.

We will identify LO1-6 and use TMS to assess their causal role in spatial vision. This requires that we measure the way TMS changes visual performance in spatial visual tasks. We will determine whether orientation and shape task performance depends crucially on processing in LO1 and LO2, but extend our previous work to see whether this remains the case when orientation and shape are defined by different attributes such as colour and motion. To extend further our assessment of the roles of LO maps in spatial vision we will consider spatial processing that requires integration of information across large spatial scales, which is required for us to assemble coherent cues that are crucial for perception of contours and symmetry. The TMS experiments will be complemented by fMRI experiments that can examine responses at a finer spatial scale allowing the separate roles of the smaller LO maps to be examined closely. Finally, we will use measures that can assess the pattern of fMRI activity to predict the type of stimulus the participant viewed which will allow us to derive new insight about how visual information is represented in the brain and how it interrelates to our perception of everyday objects.

The lateral occipital area (LO) of the human brain has been commonly identified as 'object-selective', with studies reporting larger responses in this area following the presentation of common objects compared to their scrambled counterparts. Recently however, a number of independent laboratories have confirmed the existence of multiple separate visual field maps within LO. The identification of multiple visual field maps within the larger 'object-selective' LO, offers the intriguing possibility that the object-selectivity observed in LO, emerges from the pattern of low-level analyses performed independently by these retinotopic subdivisions. Our previous work highlighted a double dissociation between two of these field maps, LO1 and LO2. LO1 was found to be causally involved in orientation processing, whereas LO2 played a crucial role in shape processing. The functional properties of the remaining LO visual field maps (LO3-6) and the potentially independent roles they play in human visual function are also currently unknown.

Using a combination of neuroimaging, neurostimulation and computational techniques, we will examine the functional roles visual field maps LO1-6 play in spatial vision. In functional magnetic resonance imaging (fMRI) experiments we will identify the six LO visual field maps in individual subjects using retinotopic mapping techniques. We will then: (1) assess the contributions of each visual field map using fMRI experiments in which we assess responses to different visual stimuli, (2) independently target these visual field maps for disruption using transcranial magnetic stimulation (TMS) while participants perform discriminations of visual stimuli and (3) determine whether signals in LO maps are sufficient to decode visual stimuli. Our multidisciplinary approach will allow us to reach a new understanding of how multiple extrastriate visual field maps contribute to perception of complex visual forms.

Who are the beneficiaries of the proposed research and how will they benefit?

The proposal constitutes basic research that offers to provide new insights into the functional properties of different visual areas in the human brain. These are important contemporary issues that will benefit researchers from a range of disciplines. The immediate beneficiaries will include researchers spanning the fields of: perceptual psychology, visual psychophysics, visual neuroscience, neuropsychology, computational vision, as well as behavioural and cognitive neuroscience.

The conceptual framework for visual processing in the brain that we are considering is a useful construct for understanding the function of the brain. This has educational value in itself and as our research progresses, we expect that undergraduate and postgraduate students will benefit from the dissemination of our findings. University students will therefore gain an enhanced educational experience.

In addition to the academic beneficiaries, one of the major benefits of this research to wider society lies in the fact that we can directly investigate human brain function without recourse to the use of animal experimentation, which is frequently not received positively by the public. Previously, establishing what causal roles specific brain areas play in perception has been achieved by undertaking lesion studies in animals, most relevantly monkeys. However, it is now accepted that transcranial magnetic stimulation, the technique we will use, can be safely used in man to cause temporary lesions. Therefore our methods offer the advantage that we can learn directly about brain function and visual processing without the need to resort to animal models. Also relevant to this is the specific region of the brain that we, which to investigate, appears to have no direct homologue in animal models.

Scientific research relating to the structure and function of the human brain interests the public and media. Therefore our work provides a good vehicle for engaging the public with science and for fostering interest in science amongst young people, helping to promote better knowledge and understanding of neuroscience amongst the wider community.

There is also potential for commercial enterprise to benefit from our research by taking our results and using them in biologically inspired pattern recognition algorithms.