Differential magnetic resonance signal change in human sensorimotor cortex to finger movements of different rate of the dominant and subdominant hand

Abstract

Functional magnetic resonance tomography (fMRI) analysis of unimanual and bimanual sequential movements in righthanders showed the following effects. First, a rate-dependent activation of the somato-motor cortex was confirmed, with faster movement rates producing higher activation both in terms of signal intensity and number of activated voxels. Second, the right hemisphere showed more activation than the left hemisphere during unimanual tasks. Third, during bimanual movements, the left hemisphere showed greater activation than the right hemisphere. Finally, while the left hemisphere showed a marked change in activation patterns from unimanual to bimanual task, the right hemisphere activation patterns were not sensitive to task changes. The hemispheric asymmetries suggest substantial left hemisphere involvement in the coordination of bimanual tasks.

Introduction

Traditional electrophysiological and current imaging techniques such as positron emission tomography (PET) and functional magnetic resonance tomography (fMRI) rely on changes in neuronal activity to draw inferences about the functional states of individual neurons and small neuron collectives in the former case and activation of relatively large collectives of neurons in the other. Simple motor tasks lend themselves well to an electrophysiological analysis, as exemplified by the correlation between neuronal changes and motor output parameters [16]. On a more global scale, PET and fMRI studies have shown a mass activational effect in the motor cortex in which higher rates of finger tapping are reflected in stronger motor cortex signals, both in terms of area of activity and strength of activity 6, 26, 30, 31, 33, 34. In case of the Schlaug et al. [31]study it could be shown that higher tapping rates lead to a stronger signal in the finger region of the primary motor cortex, and that at higher but not at lower tapping rates an additional activation in the supplementary motor area (SMA) could be seen. This contrasts with the finding of Sadato et al. [30]who reported activation of the SMA at very slow rates, but not at the fast rates, and illustrates that even in these quite simple paradigms, relatively small procedural differences can give rise to puzzling discrepancies. Nevertheless, the main effect of a rate dependent primary motor cortex response in these studies is in good agreement with the electrophysiological demonstration that movement rate is reflected in discharge rates 4, 32. There are studies in which such a main effect of rate on MRI response has not been shown 5, 29, 37and this discrepancy has to be accounted for. One possible source for the difference lies in the failure to correct for forehead motion artifacts. When movement is not controlled for, the signal/noise ratio can deteriorate to a point where significant effects can no longer be detected. Lack of movement control does not necessarily eliminate the rate effect [26], presumably because a large number of trials may still allow detection of the experimental effect, but it is clear that studies lacking control for movement artifacts need to be examined with particular care. A further source of the failure to document the rate effect lies in the number of slices that are used; small numbers of slices will not be able to deal adequately with individual differences in size and shape of the sensorimotor hand area, as discussed by White et al. [38], and may miss the area altogether.

One objective of the present study was to provide additional information about the rate dependent activation effect for simple finger movement. However, the primary purpose of the study was to examine the activation effects for the preferred and nonpreferred hands in single and bimanual hand movements. This is of interest for two reasons. First, a number of studies have suggested that there is an anatomically different organization for the sensorimotor cortex of the preferred and nonpreferred hand in monkeys [20]and humans 1, 10, 25, 39. The demonstrations are not without contradiction [38]but are of interest because of the well documented hand performance asymmetries 22, 23. Thus, our question was whether the activation patterns of the primary motor cortex for the nonpreferred left hand of righthanders would show different activation patterns than those for the preferred right hand, given similar motor tasks. This question was examined in the context of both single and bimanual motor tasks. The second point of interest was to see how cortical activation effects would vary as a function of uni- or bimanual movement. While handedness research focuses largely on single hand movement control, theoretical arguments exist which emphasize the functional requirements of bimanual integration in the genesis—both ontogenetically and phylogenetically—of handedness. The assumption is that when both hands are concurrently active, there is a sharpening of the asymmetry in hemispheric contributions to hand control [21].