Prismatic Adaptation Modulates Oscillatory EEG Correlates of Motor Preparation but Not Visual Attention in Healthy Participants

Abstract

Prismatic adaption (PA) has been proposed as a tool to induce neural plasticity and is used to help neglect rehabilitation. It leads to a recalibration of visuomotor coordination during pointing as well as to aftereffects on a number of sensorimotor and attention tasks, but whether these effects originate at a motor or attentional level remains a matter of debate. Our aim was to further characterize PA aftereffects by using an approach that allows distinguishing between effects on attentional and motor processes. We recorded EEG in healthy human participants (9 females and 7 males) while performing a new double step, anticipatory attention/motor preparation paradigm before and after adaptation to rightward-shifting prisms, with neutral lenses as a control. We then examined PA aftereffects through changes in known oscillatory EEG signatures of spatial attention orienting and motor preparation in the alpha and beta frequency bands. Our results were twofold. First, we found PA to rightward-shifting prisms to selectively affect EEG signatures of motor but not attentional processes. More specifically, PA modulated preparatory motor EEG activity over central electrodes in the right hemisphere, contralateral to the PA-induced, compensatory leftward shift in pointing movements. No effects were found on EEG signatures of spatial attention orienting over occipitoparietal sites. Second, we found the PA effect on preparatory motor EEG activity to dominate in the beta frequency band. We conclude that changes to intentional visuomotor, rather than attentional visuospatial, processes underlie the PA aftereffect of rightward-deviating prisms in healthy participants.SIGNIFICANCE STATEMENT Prismatic adaptation (PA) has been proposed as a tool to induce neural plasticity in both healthy participants and patients, due to its aftereffect impacting on a number of visuospatial and visuomotor functions. However, the neural mechanisms underlying PA aftereffects are poorly understood as only little neuroimaging evidence is available. Here, we examined, for the first time, the origin of PA aftereffects studying oscillatory brain activity. Our results show a selective modulation of preparatory motor activity following PA in healthy participants but no effect on attention-related activity. This provides novel insight into the PA aftereffect in the healthy brain and may help to inform interventions in neglect patients.

Keywords: EEG; aftereffects; attention orienting; brain oscillations; motor preparation; prismatic adaptation.

Figure 1.

Experimental setup and paradigm. A, Experimental timeline. B, Experimental paradigm. Each trial started with a fixation cross, followed by an attentional cue (the bottom left or right section of the central rhombus turning green) instructing participants to covertly attend to the left or right bottom visual field placeholder. After 1500 ms, a second, motor preparation cue (big or small triangle) appeared in the left or right placeholder (80% at attended and 20% at unattended position) pointing either to the left or to the right (probability of 50%). The motor preparation cue indicated which response (left or right hand) the participants needed to prepare. After another 1500 ms, a go signal (green vertical line) instructed participants to perform the prepared action. EEG was analyzed in terms of oscillatory alpha and beta activity in the two 1500 ms Post-Cue intervals, covering anticipatory attention and preparatory motor processes to the left or right side of space, respectively, as well as in terms of visual evoked potentials to the motor cue (also serving as visual target).

Figure 2.

PA setup and timeline. Participants point to targets on a curved, transparent panel. Preexposure (prismatic goggles off) involves pointing in free viewing conditions (both pointing movements and targets visible) followed by occluded (blinded) pointing to visible targets. Participants were then asked to wear the googles (rightward orientation or neutral lenses) during free viewing pointing (exposure, goggles on). Adaptation is then tested immediately after exposure with blinded pointing to targets (aftereffect).

Figure 3.

PA pointing displacement. Mean pointing displacement (expressed in degrees of visual angle) throughout the PA procedure (preexposure free viewing/preexposure blinded, early and late exposure, aftereffect) is plotted for each condition. Solid line indicates pointing when wearing real (prismatic) lenses (prismatic goggles). Dotted line indicates pointing with neutral lenses (neutral goggles). Negative values indicate a leftward-pointing displacement. Positive values indicate a rightward displacement. Error bars indicate SEM (standard error of the mean). *p < 0.001, significant difference between conditions.

Figure 4.

Alpha modulation by attention orienting. A, Time-frequency representations of the anticipatory attention-related alpha modulation are shown separately across rows for each PA condition (Pre-/Post-Prism, Pre-/Post-Neutral) for two posterior EOIs (left and right columns) by contrasting attention right and attention left trials [(Power_{Attention right} − Power_{Attention left})/common denominator]. The electrodes included in the left and right EOIs are indicated by black dots in the central maps (P3/4, P5/6, P7/8, PO3/4, PO7/8, and O1/2). Middle column represents the topography of alpha modulation (8–12 Hz) between 0.2 and 1 s after attentional cue onset (black rectangle). B, Cluster-based analysis. Difference maps of alpha modulation between conditions (8–12 Hz, 0.2–1 s Post-Cue). Raw effects are shown for each simple comparison on the left (Pre- vs Post-Prism; Pre- vs Post-Neutral) and for the Exposure × PA interaction on the right. No significant differences were identified by cluster-based statistics (all p values > 0.05). C, EOI analysis. AMI [AMI = (Power_{Attention Contra} − Power_{Attention Ipsi})/average over all conditions] in the alpha band (8–12 Hz, 0.2–1 s) over posterior sites (P3/4, P5/6, P7/8, PO3/4, PO7/8, and O1/2). Statistical analysis revealed no significant 2 × 2 interactions. Error bars indicate SEM.

Figure 5.

Alpha/mu modulation by motor preparation. A, Time-frequency representations of the motor preparation-related alpha/mu modulation are shown separately across rows for each PA condition (Pre-/Post-Prism, Pre-/Post-Neutral) for two central EOIs (left and right columns) by contrasting right- and left-hand motor preparation trials [(Power_{Right Hand} − Power_{Left Hand})/common denominator]. The electrodes included in the left and right EOIs are indicated by black dots (C3/4, CP3/4) in the central maps. Middle column represents the topography of alpha modulation (8–12 Hz) between 0.5 and 1.2 s after motor cue onset (black rectangle). B, Cluster-based analysis. Difference maps of alpha modulation between conditions (8–12 Hz). Raw effects are shown for each simple comparison on the left (Pre- vs Post-Prism; Pre- vs Post-Neutral) and for the Exposure × PA interaction on the right. No significant cluster was identified (p > 0.05). C, EOI analysis. MPI [MPI = (Power_{Hand Contra} − Power_{Hand Ipsi})/average over all conditions] in the mu band (8–12 Hz, 0.5–1.2 s) over central sites (C3/4, CP3/4). Statistical analysis revealed no significant 2 × 2 interactions. Error bars indicate SEM.

Figure 6.

Beta modulation by motor preparation. A, Time-frequency representations of the motor preparation-related beta modulation are shown separately across rows for each PA condition (Pre-/Post-Prism, Pre-/Post-Neutral) for two central EOIs (left and right columns) by contrasting right- and left-hand motor preparation trials [(Power_{Right Hand} − Power_{Left Hand})/common denominator]. The electrodes included in the left and right EOIs are indicated by black dots (C3/4, CP3/4) in the central maps. Middle column represents the topography of beta modulations (16–25 Hz) between 0.5 and 1.2 s after the cue (black rectangle). B, Cluster-based analysis. Difference maps of beta modulation between conditions (16–25 Hz, 0.5–1.2 s Post-Motor cue). Raw effects are shown for each simple comparison on the left (Pre- vs Post-Prism; Pre- vs Post-Neutral) and for the Exposure × PA interaction on the right. 2 × 2 (Prism/Neutral vs Pre/Post) cluster-based permutation analyses identified a significant interaction cluster (p < 0.03, see black dots in right interaction map). Follow-up simple tests revealed a significant cluster (p = 0.008) for Pre- versus Post-Prism PA but not for Pre- versus Post-Neutral lenses (see left maps). C, EOI analysis. MPI [MPI = (Power_{Hand Contra} − Power_{Hand Ipsi})/average over all conditions] in the beta band (16–25 Hz, 0.5–1.2 s) over central sites (C3/4, CP3/4). Positive values indicate the expected, contralateral versus ipsilateral modulation. Statistical analysis revealed a significant interaction of Exposure × Time × Hemisphere (p < 0.05). The MPI over the right hemisphere increased Post-PA (p = 0.015). Error bars indicate SEM. **p < 0.05.

Figure 7.

ERPs to targets/motor cues. A, P1. B, N1 amplitudes and latencies before and after PA (Prism condition on the left and Neutral control on the right) are shown separately for hemispheres (Ipsilateral and Contralateral to the target position), validity of attentional cueing (Valid and Invalid), and target position (Left and Right). Anticipatory attention modulated the amplitude and latency of the P1 and N1 components independently of PA. Electrodes: PO7/8.