Elsevier

Neuropsychologia

Volume 48, Issue 9, July 2010, Pages 2595-2601
Neuropsychologia

Neural mechanisms underlying spatial realignment during adaptation to optical wedge prisms

https://doi.org/10.1016/j.neuropsychologia.2010.05.006Get rights and content

Abstract

Visuomotor adaptation to a shift in visual input produced by prismatic lenses is an example of dynamic sensory-motor plasticity within the brain. Prism adaptation is readily induced in healthy individuals, and is thought to reflect the brain's ability to compensate for drifts in spatial calibration between different sensory systems. The neural correlate of this form of functional plasticity is largely unknown, although current models predict the involvement of parieto-cerebellar circuits. Recent studies that have employed event-related functional magnetic resonance imaging (fMRI) to identify brain regions associated with prism adaptation have discovered patterns of parietal and cerebellar modulation as participants corrected their visuomotor errors during the early part of adaptation. However, the role of these regions in the later stage of adaptation, when ‘spatial realignment’ or true adaptation is predicted to occur, remains unclear. Here, we used fMRI to quantify the distinctive patterns of parieto-cerebellar activity as visuomotor adaptation develops. We directly contrasted activation patterns during the initial error correction phase of visuomotor adaptation with that during the later spatial realignment phase, and found significant recruitment of the parieto-cerebellar network – with activations in the right inferior parietal lobe and the right posterior cerebellum. These findings provide the first evidence of both cerebellar and parietal involvement during the spatial realignment phase of prism adaptation.

Introduction

When participants point at targets under visual guidance while wearing prism lenses that displace the visual field laterally, they demonstrate errors in pointing that are in the direction of the optical displacement. These errors rapidly decrease with repeated pointing, as the participant compensates for the misalignment between visual inputs and the felt position of the hand. When the prism lenses are removed, pointing errors are again apparent, but this time in the direction opposite to that of the prism-induced visual displacement, indicating an adaptive shift in perceived hand position. This aftereffect reflects underlying neural adaptation to the visuomotor disparity (Harris, 1965, Held, 1965, Redding et al., 2005).

In studies of prism adaptation, two distinct phases have been identified: error correction and spatial realignment (Redding and Wallace, 1993, Redding et al., 2005). Error correction, or strategic control, describes a process of error reduction in movement plans whereby the participant anticipates an error and plans a movement that minimises the perturbation (Redding and Wallace, 1993, Redding and Wallace, 1996). Although the initial pointing errors associated with prism adaptation are reduced to zero within the first few reaches, a further period of repeated pointing is essential for the aftereffect (and spatial realignment) to develop fully (Redding & Wallace, 1993). In fact, strategic error correction (as indexed by the increasing accuracy of early pointing trials) and spatial realignment (as indexed by the aftereffect) are thought to be two interrelated but independent processes (Fernandez-Ruiz et al., 2007, Michel et al., 2007, Newport and Jackson, 2006, Redding and Wallace, 1993, Redding and Wallace, 1996).

The neural circuitry underlying each phase of adaptation remains unclear. Human lesion data indicate a role for the cerebellum in both spatial realignment and error correction; either process can be independently impaired, depending on the regions damaged within this structure (Fernandez-Ruiz et al., 2007, Martin et al., 1996, Pisella et al., 2005). The parietal lobe also appears to be involved in the strategic error correction phase of adaptation, but its role in spatial realignment is somewhat contentious. Bilateral lesions of the parietal lobe can impair the ability to strategically correct for errors induced by prismatic displacement (Newport and Jackson, 2006, Pisella et al., 2004) while still enabling the development of an aftereffect (although see Newport & Jackson, 2006). However, both bilateral and unilateral parietal lesions often produce abnormally large aftereffects of prism adaptation and generalization to other, untrained, spatial behaviours (Pisella et al., 2004, Sarri et al., 2008), implying parietal involvement in at least modulating the degree of spatial realignment (Michel, 2006). Based on such lesion data, current models of prism adaptation propose a network of parietal and cerebellar regions involved in detecting the visual-motor misalignment produced by prisms and invoking changes to internal models of movement to accommodate for this (e.g., Michel, 2006, Newport and Jackson, 2006, Pisella et al., 2006, Ramnani, 2006, Redding and Wallace, 2006, Serino et al., 2006). What does this mean for the intact brain? Findings from the few available brain imaging studies of prism adaptation support a role for parietal and cerebellar regions in the initial error correction phase (Clower et al., 1996, Danckert et al., 2008, Luauté et al., 2009). There have been no imaging studies, however, that have clearly identified neural mechanisms distinctive to realignment. Identifying the neural correlates of this later phase is critical for understanding the therapeutic effects of prism adaptation in parietal patients (e.g., Pisella et al., 2006) as well as the mechanisms behind short-term plasticity of spatial maps in the normal brain. Thus the primary goal of the present study was to examine the neural basis of prism-induced spatial realignment in the healthy brain.

Several studies have examined neural activation patterns associated with the initial error correction phase of prism adaptation (Clower et al., 1996, Danckert et al., 2008, Luauté et al., 2009). Together, their findings strongly implicate the inferior parietal cortex, contralateral to the pointing arm, in the process of detecting and consciously correcting for the prism-induced misalignment between vision and perceived hand position. In their pioneering study, Clower et al. (1996) measured changes in regional cerebral blood flow (rCBF) associated with pointing movements under the influence of laterally displacing prisms. In their study, the direction of displacement of the prism lenses was reversed every five trials, thus forcing the participant to continually correct for the prism-induced errors in pointing. Under these conditions, activation within the left intraparietal sulcus (IPS) was associated with error correction during prism adaptation. More recently, Danckert et al. (2008) used fMRI to investigate the neural basis of error correction over the course of the first ten pointing trials of prism adaptation. Their findings confirmed a role for the left anterior intraparietal sulcus during the first three trials of prism exposure when participants were experiencing large pointing errors, relative to later trials on which errors had been reduced. In addition, increased activity in regions within the anterior cingulate and primary motor cortex were also found to be associated with these early error correction trials. Further evidence for the role of the parietal lobe in correcting for prism-induced errors comes from a recent event-related fMRI study in which trial-by-trial decreases in the magnitude of the BOLD response within the left anterior IPS closely matched participants’ behavioural reduction in pointing errors (Luauté et al., 2009). Luauté et al. (2009) also reported a trial-by-trial increase in activation within the posterior occipital sulcus (POS) over the same period of error reduction. The authors suggested that these two sites, IPS and POS, might be involved in error detection and correction, respectively.

Unlike previous studies, Luauté et al. (2009) allowed participants to continue pointing well beyond the initial error correction phase, thus imaging the entire adaptation process (including spatial realignment) for the first time. Crucially, the only sub-threshold activation that the authors could attribute to the later, spatial realignment period (and thus to development of the aftereffect) was a gradual, but subtle increase in cerebellar activity that outlasted the initial period of error correction. The trial-by-trial analysis is an elegant approach for exploring brain activation associated with a temporally unfolding process such as adaptation. However, unlike error correction, spatial realignment is not associated with an observable behaviour during the course of adaptation, and neither its point of initiation nor its rate of development are known. In this context, the slight increase in cerebellar activation observed by Luauté et al. (2009) could have multiple possible causes, including slow, random variations in the hemodynamic signal, as the authors themselves acknowledge.

The lack of clear behavioural markers for spatial realignment poses methodological complexities for isolating activation patterns related to realignment. In fact, error correction and realignment are not necessarily sequential processes, and likely overlap in time (Redding and Wallace, 1993, Redding et al., 2005). Processes relating to realignment are nevertheless predicted to dominate during the later trials of adaptation (Redding and Wallace, 1993, Redding et al., 2005). In the present study, we employed a blocked fMRI design to directly compare brain activation elicited during the initial half of adaptation trials (including the error correction phase) with that during the later half of trials (corresponding principally to spatial realignment). Blocking the adaptation session in this manner allowed us to contrast trials predominantly associated with realignment against the earlier trials in which minimal realignment was expected to have occurred. By subtracting neural responses for error correction from those associated with spatial realignment, and assuming an overlapping transition from one phase to the next, this approach provides a conservative estimate of the activation unique to spatial realignment. Crucially, our approach yielded robust neural responses during the spatial realignment phase, which included subregions of the cerebellum and the inferior parietal cortex.

Section snippets

Participants

Fourteen right-handed male participants (20–42 years) with no history of neurological disease participated. All participants gave informed consent. One participant's data set was removed due to large head movements during scanning (>5 mm), and another data set was removed due to a technical error during data collection.

Materials and procedure

To enable participants to easily alternate between prismatic and normal vision within the scanner, they wore plastic goggles modified to include a 10° wedge-based optical prism

Behavioural data

As apparent from Fig. 2, participants initially displayed large leftward errors in pointing during the error correction phase of prism adaptation. These errors rapidly reduced to zero by the end of the error correction phase and pointing accuracy was maintained throughout the ensuing spatial realignment phase. This was confirmed statistically. We conducted a two-way repeated measures analysis of variance on pointing errors (distance between pointing response and target) obtained during the

Discussion

Consistent with current models of adaptation (Newport and Jackson, 2006, Pisella et al., 2006, Ramnani, 2006), and with the fMRI findings to date (Clower et al., 1996, Danckert et al., 2008, Luauté et al., 2009), our results provide evidence for the involvement of parietal and cerebellar structures during prism adaptation. Importantly, our findings show that neural responses in the right cerebellum and inferior parietal lobe are recruited during the later, spatial realignment phase of prism

Conclusion

In summary, we have provided evidence for both cerebellar and inferior parietal lobe activity during the spatial realignment phase of prism adaptation in normal, healthy adults. These findings have implications for our understanding of the neural mechanisms underlying visuomotor adaptation in the normal brain, the modulation of adaptive aftereffects, and their translation to spatial cognition.

Acknowledgements

This research was supported by grants from ANZ Trustees (H.C. and J.B.M) and the National Health and Medical Research Council (Australia) (J.B.M. and R.E.).

References (39)

  • D.M. Clower et al.

    Role of posterior parietal cortex in the recalibration of visually guided reaching

    Nature

    (1996)
  • D.M. Clower et al.

    The inferior parietal lobule is the target of output from the superior colliculus, hippocampus, and cerebellum

    The Journal of Neuroscience

    (2001)
  • C. Colent et al.

    Cognitive bias induced by visuo-motor adaptation to prisms: A simulation of unilateral neglect in normal individuals?

    Neuroreport

    (2000)
  • J. Danckert et al.

    Direct effects of prismatic lenses on visuomotor control: An event-related functional MRI study

    European Journal of Neuroscience

    (2008)
  • V. Della-Maggiore et al.

    Stimulation of the posterior parietal cortex interferes with arm trajectory adjustments during the learning of new dynamics

    Journal of Neuroscience

    (2004)
  • M. Desmurget et al.

    Role of the posterior parietal cortex in updating reaching movements to a visual target

    Nature

    (1999)
  • C. Farrer et al.

    The angular gyrus computes action awareness representations

    Cerebral Cortex

    (2008)
  • C.S. Harris

    Perceptual adaptation to inverted, reversed and displaced vision

    Psychological Review

    (1965)
  • R. Held

    Plasticity in sensory-motor systems

    Scientific American

    (1965)
  • Cited by (113)

    View all citing articles on Scopus
    View full text