How do the two halves of our brain work together?

Our brain is divided into left and right hemispheres, each with its own unique qualities, but much of our cognition depends on processing that spans the two hemispheres. How does this occur? One way to find out is to investigate what happens when the connections between the two hemispheres are disrupted in a 'split brain'. A split brain (Roser and Gazzaniga 2004; 2006; 2009) results from cutting the corpus callosum, a large fibre tract that connects the hemispheres, to relieve epilepsy. This means that the two halves of the brain can be tested in relative isolation compared to those in a normally-connected brain.

Testing split-brain patients, and people with a congenital lack of the corpus callosum, we have shown (Roser and Corballis, 2002; 2003; Roser et al., 2012) that the two hemispheres in the normal brain co-operate extensively, even when carrying out the simplest of tasks such as pressing a button in response to a light. Accumulating evidence suggests that there are real differences between the two hemispheres, but that most processes involve a complex interplay between the two that depends on the corpus callosum.

To better understand how cognition depends on interhemispheric connections in the neurologically-normal brain, and how these connections are affected by age, we are currently running studies with young and aged participants using functional MRI and diffusion-tensor imaging (DTI). (See the section labelled 'current'). We (Linnet & Roser, 2012) have found that normal ageing produces some symptoms of hemispheric disconnection that manifest in motor responses to perceptual stimuli.

We have also investigated brain connectivity in the presence of callosal curvilinear lipoma, which are congenital malformations that occur as a result of the persistence and maldifferentiation of the primitive meninges during embryonic development, and have been associated with callosal malformation ranging from hypogenesis to agenesis. We investigated hemispheric interaction in a man with a peri-splenial lipoma using diffusion-tensor imaging, comparison to his neurologically-normal identical twin, and comparison to a control group. Reduced structural connectivity through the posterior corpus callosum was associated with an abnormal pattern of interhemispheric interaction, highlighting the importance of this region, and suggesting that lipomas are associated with subtle abnormalities in brain and behaviour even when white matter appears grossly normal.

(See Roser, M. E., Corballis, M. C., Jansari, A., Fulford, J., Benattayallah, A., & Adams, W. M. (2012). Bilateral redundancy gain and callosal integrity in a man with callosal lipoma: A diffusion-tensor imaging study. Neurocase, 18(3), 185-198.)



How do we understand the world around us?

Testing split-brain patients, and using brain imaging, can also tell us about how the two halves of the brain contribute to our understanding of the world. Our research suggests that what we think of as quite sophisticated concepts, such as causality, have their roots in low-level perceptual processes. Furthermore, the two cerebral hemispheres make different contributions to understanding the physical world. In split-brain patients we (Roser, Fugelsang, Dunbar, Corballis, and Gazzaniga, 2005) found that the right hemisphere carries out perceptual processes that are essential for our ability to 'see' causality in the interaction of objects in the world around us, while the left hemisphere is essential for inferential processes that extend our understanding of causality beyond our immediate physical experience. Using functional magnetic resonance imaging (fMRI) we (Fugelsang, Roser, Corballis, Gazzaniga, and Dunbar, 2005) found that perceiving causality in object interactions involved activity in a network of frontal and parietal brain regions in the right hemisphere, supporting our findings in split-brain patients.

We have further explored the role of low-level processes in understanding causality using event-related potentials (ERPs). By manipulating the kinematic properties of collision events we (Roser, Fugelsang, Handy, Dunbar, and Gazzaniga, 2009) found that brain electrical activity elicited by 'oddball' events was modulated by the physical plausibility of object interactions. This finding is consistent with theoretical accounts that posit that we understand the world by mapping stimulus input on to pre-existing schemas, derived from experience, that incorporate information about everyday physical objects and our actions on them.

We have further investigated the contributions of the two cerebral hemispheres to understanding variation in the world by studying incidental learning of visual statistics in the split brain (Roser, Fiser, Aslin, and Gazzaniga, 2011). We found that only the right hemisphere was sensitive to simple patterns of covariation in arrays of visual objects. This finding points to the importance of perceptual processes supported by the right hemisphere in allowing us to navigate and make sense of a world that is constantly changing. Testing in the split brain also allowed us to better understand visual statistical learning. Our results were consistent with a model in which the early stages of learning are supported by bottom-up processes which are later integrated with top-down knowledge. Our recent work (Karuza, Emberson, Roser, Gazzaniga, Cole, Aslin, and Fiser, 2014) using fMRI has shown that learning is not fully captured by a single, fixed “learning” network, but is reflected at least partially in dynamic shifts in functional connectivity across numerous cortical and subcortical areas.