A chronopsychophysiological study of mental rotation

Abstract

The cognitive task of imagining how an object will Iook if rotated from the orientation in which it appears commonly results in linearly increasing reaction times proportional to the angular displacement of the required rotation. Typically subjects are presented with pictures of two objects simultaneously and required to determine if the items are the same or different. In performing this task subjects seem to imagine moving one of the objects in a continuous manner until it matches the other, in a similar manner to actually performing the rotation manually. This phenomenon has become known as "mental rotation". A great many studies have been performed confirming this finding. With the increasing availability of neuroimaging technology there has been a burgeoning interest in the underlying neural basis of mental rotation. While the parietal cortex has consistently been shown to be involved in solving these cognitive tasks, evidence for the recruitment of other brain regions has been considerably more variable. In particular, there is extensive debate concerning the nature of any possible involvement of the motor cortices. If such areas are involved with the mental rotation process proper it suggests that such activity may involve visuomotor as well as Visuospatial processing. This thesis presents evidence from four consecutive studies concerning the timecourse of neural events and the underlying neural generators associated with mental rotation tasks. The data presented were gathered using a combination of event-related potentials (ERPs) and electroencephalography (EEG) spectral analyses. In each study, pictures of Ieft and right hands, and nonsense objects in one instance, were used as stimuli. The hand may be a special class of object when it comes to mental rotation as it seems to invoke an imagined rotation of the subject's own hand. It was reasoned that using pictures of hands was most likely to engage motor imagery. The aim of the first study was to establish that there were reliable and reproducible ERPs associated with the mental rotation of hands. The results revealed consistent and task correlated ERPs; involving a progression of latencies and brain regions starting with early activations in the occipital lobe, followed by involvement of central parietal regions and later activations in central and frontal areas. Study 2 investigated the role of central motor structures using ERPs. It revealed that mental rotation was most strongly associated with ERP amplitude modulations during a time window of 600-800 ms. Furthermore, this amplitude modulation varied systematically with the orientation of the presented stimulus. The neural generators specific to mental hand rotation were suggested to be in extra-striate and parietal cortices. Study 3 characterised the EEG spectral response to mental hand rotation. A go/no go task separated motor processes associated with mental rotation from those underlying the behavioural response. Beta band fluctuations over parietal regions related directly to the mental rotation task, while beta activity at central electrodes characterised the preparation for and emission of the behavioural response. Similarly, sensorimotor mu rhythm was associated with subjects' overt response. In the final study stimuli included pictures of Ieft and right hands as well as inanimate nonsense objects. The aim was to assess the proposal that different neural substrates underlie object-based and bodysegment- based mental rotations. It was hypothesised that orientation-dependent ERP components would exhibit different stimulus-dependent topographical distributions. The results of the study did not support this hypothesis and demonstrated no significant topographical differences between the stimuli. The data did suggest however, that in the case of mental hand rotation the task was more difficult and the neural generators recruited a larger neural population, as reflected by an increase in ERP amplitudes. In conclusion these studies confirmed a dominant role for parietal cortices in the mental rotation of hands. They suggested that motor regions are involved in movement anticipation and the overt behavioural response. Furthermore, the results demonstrated that anterior motor cortices are not directly involved with neural processes required to make the spatial transformation associated with mental rotation tasks.