The Neural Bases of Egocentric Spatial Representation for Extracorporeal and Corporeal Tasks: An fMRI Study

(1) Background: Humans use reference frames to elaborate the spatial representations needed for all space-oriented behaviors such as postural control, walking, or grasping. We investigated the neural bases of two egocentric tasks: the extracorporeal subjective straight-ahead task (SSA) and the corporeal subjective longitudinal body plane task (SLB) in healthy participants using functional magnetic resonance imaging (fMRI). This work was an ancillary part of a study involving stroke patients. (2) Methods: Seventeen healthy participants underwent a 3T fMRI examination. During the SSA, participants had to divide the extracorporeal space into two equal parts. During the SLB, they had to divide their body along the midsagittal plane. (3) Results: Both tasks elicited a parieto-occipital network encompassing the superior and inferior parietal lobules and lateral occipital cortex, with a right hemispheric dominance. Additionally, the SLB > SSA contrast revealed activations of the left angular and premotor cortices. These areas, involved in attention and motor imagery suggest a greater complexity of corporeal processes engaging body representation. (4) Conclusions: This was the first fMRI study to explore the SLB-related activity and its complementarity with the SSA. Our results pave the way for the exploration of spatial cognitive impairment in patients.


Introduction
Multiple sources of afferent information (from vestibular, visual, and proprioceptive receptors) and efferent information (from motor effectors) are needed to elaborate internal spatial representations of the body in space [1,2]. These inputs can be integrated into different and changing systems of coordinates, depending on the person's position in space and the nature of the task. The egocentric reference frame encodes spatial information centered on body coordinates [3][4][5][6][7][8]. This frame involves two spatial components, depending on the intended action: extracorporeal and corporeal [9]. The extracorporeal component enables individuals to locate stimuli perceived in the extrapersonal space relative to their body and orient their actions in the environment. The corporeal component allows individuals to elaborate personal representations of the location and configuration of their body centered in space, in order to perform tasks such as personal care [4,10]. Two tasks have been used to explore these egocentric reference frames: the subjective straight-ahead task (SSA) and, more recently, the subjective longitudinal body plane task (SLB). The SSA explores extracorporeal space perception centered on the body. Participants imagine a plane starting from their midsagittal body and going straight ahead, dividing the extracorporeal space into two parts [4,10,11]. In the SLB, participants are asked to divide their corporeal space into two parts. They indicate with a button press when the bar virtually divided their corporeal space into two equal parts along their body's midsagittal plane [5,[11][12][13][14]. There are well-known clinical differences between extracorporeal and corporeal tasks [5,12], and the integration of relevant information can be differentially affected following brain damage [15]. For example, differences between extracorporeal and corporeal abilities have been identified when assessing unilateral spatial neglect. An impact of body representation in space on asymmetrical postural behaviors has also been reported. One previous study highlighted a strong relationship between SLB disturbance and postural asymmetry in patients with chronic stroke [5]. However, the neural basis of these differences has never been investigated. It is essential to improve knowledge in this field, as gaining a better understanding of the mechanisms involved in elaborating the egocentric reference frame would inform rehabilitation strategies for patients with spatial cognition disorders following stroke.
The present work was the first to use fMRI to study the commonalities and differences in brain activity between two egocentric tasks involving either extracorporeal or corporeal space. We hypothesized that both tasks would activate a bilateral fronto-parieto-occipital network, especially the posterior parietal and premotor cortices, with right hemispheric predominance [15][16][17][18]. Moreover, we expected the two tasks to activate distinct subregions of the network. This research among healthy participants was an ancillary study to a multicenter study exploring spatial cognitive disorders in patients with stroke.

Participants
We included 17 healthy volunteers (7 men and 10 women: Mean age = 50.4 years, SD = 14.9, range 23-70). They were all right handed on their declaration. All participants gave their written informed consent to take part in this study, which was approved by the Poitiers-France III West institutional review board (no. 05/12/16; AVCPOSTIM study-clinical trial identifier NCT01677091). Exclusion criteria were (1) contraindication to magnetic field exposure, (2) neurological disease, and (3) history of balance disorders. All participants had normal or corrected-to-normal vision.

Experimental Design
The SSA and SLB tasks were performed in a supine position in an MRI scanner. For the SSA task, participants were instructed to divide the extracorporeal space in front of them into two halves [5,10,12,19]. A vertical green bar scanned the black screen from left to right and from right to left, with different random starting positions within the range of −7 • to +7 • , at a speed of 2 • per second [20]. The instruction was "Press the button when the bar is straight ahead and divides the space in front of you into two equal parts" (Figure 1 and Supplementary File S1). During the SLB task, a green bar moved along the plane orthogonal to the line going through the navel [5,13,21]. The bar rotated alternately leftward and rightward, with different random starting angles within the range of −30°/+30°, at a speed of 2° per second. The instruction was "Press the button when the bar is aligned with a virtual plane dividing your corporeal space into two equal parts" (Figure 2 and Supplementary File S2). During the SLB task, a green bar moved along the plane orthogonal to the line going through the navel [5,13,21]. The bar rotated alternately leftward and rightward, with different random starting angles within the range of −30 • /+30 • , at a speed of 2 • per  During the SLB task, a green bar moved along the plane orthogonal to the line going through the navel [5,13,21]. The bar rotated alternately leftward and rightward, with different random starting angles within the range of −30°/+30°, at a speed of 2° per second. The instruction was "Press the button when the bar is aligned with a virtual plane dividing your corporeal space into two equal parts" (Figure 2 and Supplementary File S2). During the control condition, the same sequence of bar movements was presented, but participants were instructed to press the button when the bar changed direction.
Each fMRI sequence lasted 3 min and followed a block design: 30 s blocks of task and 30 s block of control conditions. In each block, the bar passed through the center six times at a constant speed of two degrees per second [20]. The process is similar for both tasks. The tasks were implemented using E-Prime 2.0 Professional software (Psychology Software Tools, Pittsburgh, PA, USA) and presented using goggles (NordicNeuroLab, Bergen, Norway) attached to the head coil to immerse participants in the task without any spatial clues. The instruction was presented via the goggles just before the task, in different colors for task and rest at the beginning of each block. To perform the tasks, participants reported their perception by pressing a button with their right index finger. Before the experiment, they practiced outside the scanner until they were able to perform the tasks. They all confirmed they understood the instructions. The difference between target identification (respectively, straight ahead or longitudinal axis) and button press time were computed both for SSA and SLB tasks.

MRI Acquisition
MRI acquisition was performed on a 3 tesla scanner (Magnetom Verio; Siemens Medical Solutions, Erlangen, Germany) with a 12-channel head coil. The alignment of head, shoulders, trunk, and limbs was checked by the experimenter before MRI acquisition. The head and the trunk were maintained with foam pads. During the control condition, the same sequence of bar movements was presented, but participants were instructed to press the button when the bar changed direction.
Each fMRI sequence lasted 3 min and followed a block design: 30 s blocks of task and 30 s block of control conditions. In each block, the bar passed through the center six times at a constant speed of two degrees per second [20]. The process is similar for both tasks. The tasks were implemented using E-Prime 2.0 Professional software (Psychology Software Tools, Pittsburgh, PA, USA) and presented using goggles (NordicNeuroLab, Bergen, Norway) attached to the head coil to immerse participants in the task without any spatial clues. The instruction was presented via the goggles just before the task, in different colors for task and rest at the beginning of each block. To perform the tasks, participants reported their perception by pressing a button with their right index finger. Before the experiment, they practiced outside the scanner until they were able to perform the tasks. They all confirmed they understood the instructions. The difference between target identification (respectively, straight ahead or longitudinal axis) and button press time were computed both for SSA and SLB tasks.

MRI Acquisition
MRI acquisition was performed on a 3 tesla scanner (Magnetom Verio; Siemens Medical Solutions, Erlangen, Germany) with a 12-channel head coil. The alignment of head, shoulders, trunk, and limbs was checked by the experimenter before MRI acquisition. The head and the trunk were maintained with foam pads.
In total, the experiment was completed in 15 min. Images were preprocessed and analyzed using Statistical Parametric Mapping (SPM12; Wellcome Department of Imaging Neuroscience, University College London, London, UK [22]; revision 7487) implemented in MATLAB. Morphological images were segmented, corrected for bias, and normalized to Montreal Neurological Institute atlas space (MNI).
Slice timing, spatial realignment, distortion correction, coregistration to the morphological images, and normalization were applied, followed by 6 mm full width at half maximum (FWHM) isotropic Gaussian kernel smoothing. The canonical hemodynamic response function and its temporal derivative were used for model estimation, and three contrasts were evaluated: "SSA > control condition", "SLB > control condition", and "SLB > SSA".

Statistical Modeling and Inference Analysis
Given that a large number of regions of interest are reported in the literature, we ran voxelwise analyses across the whole brain. A cluster extent of 25 voxels was determined by running a Monte Carlo simulation [23][24][25] using cluster threshold betas based on an individual voxel threshold of p = 0.001 and a corrected threshold of p = 0.05. The two tasks were compared using a paired t test with a cluster extent of 25 voxels.

Results
Subjects responded within the standard to both tasks showing the reliability of the tasks. The SSA mean deviation was −0.4 • , SD 1.3; the SLB mean deviation was −1 • , SD 0.3. The mean response times of both tasks were not significantly different suggesting a similar complexity (p = 1.17).

Subjective Longitudinal Body Plane Task (SLB)
The "SLB > control condition" contrast highlighted a fronto-parieto-occipital network extending to the temporal lobe and insula, with right-hemispheric predominance (75.5% of clusters located in right hemisphere). Activated clusters are shown in Figure 4 and Table 2.
Activated voxels were distributed as follows:

Subjective Longitudinal Body Plane Task (SLB)
The "SLB > control condition" contrast highlighted a fronto-parieto-occipital network extending to the temporal lobe and insula, with right-hemispheric predominance (75.5% of clusters located in right hemisphere). Activated clusters are shown in Figure 4 and Table 2.  Table 2. Activated areas highlighted by "SLB > control condition" contrast. The number of voxels is given for each cluster, together with the side (i.e., left or right hemisphere), the cluster composition (i.e., the percentage of voxels in each area assigned using AAL3 and brainnetome atlas), MNI coordinates and t value of the peak. Only activated areas covering more than 10% of the cluster are shown.

Subjective Longitudinal Body Plane Task versus Subjective Straight-Ahead Task
Subjects responded within the standard to both tasks showing the reliability of the tasks. The SSA mean deviation was −0.4 • , SD 1.3; the SLB mean deviation was −1 • , SD 0.3. The mean response times of both tasks were not significantly different suggesting a similar complexity (p = 1.17).
Both tasks elicited a parieto-occipital network that encompassed the superior and inferior parietal lobules, precuneus, and lateral occipital cortex. The "SLB > SSA" comparison revealed a larger activation area (2055 vs. 633 voxels) and higher intensity for the SLB task than for the SSA task. Moreover, the "SLB > SSA" contrast revealed activation in the left angular and precentral gyri ( Figure 5, Table 3).  Table 3. Activated areas highlighted by "SLB > SSA" contrast. The number of voxels is given for each cluster, together with the side (i.e., left or right hemisphere), the cluster composition (i.e., the percentage of voxels in each area assigned using AAL3 and brainnetome atlas), MNI coordinates and t value of the peak. Only activated areas covering more than 10% of the cluster are shown.   Table 3. Activated areas highlighted by "SLB > SSA" contrast. The number of voxels is given for each cluster, together with the side (i.e., left or right hemisphere), the cluster composition (i.e., the percentage of voxels in each area assigned using AAL3 and brainnetome atlas), MNI coordinates and t value of the peak. Only activated areas covering more than 10% of the cluster are shown.

Discussion
Using fMRI, the present study investigated the neural bases of processes elicited by two different egocentric tasks in healthy participants: SSA and SLB. This was the first time that the SLB task had been implemented in fMRI, reproducing the clinical assessment setting in a supine position without visual clues. We demonstrated its feasibility and complementarity with the SSA task [20]. These tasks can both be administered to patients with spatial cognitive disorders following stroke. Our results will enhance understanding of the egocentric reference frame concept both in healthy participants and in patients with a brain lesion.
We found strong activity in the right lateral occipital cortex implicated in visual perception, corroborating previous experimental findings [43]. As we used a moving bar instead of a static image, the motion may have increased occipital activity.
Along these lines, we found less extensive frontal activation in the superior and inferior frontal gyri than in previous studies [27,29]. However, these studies compared an egocentric task with an allocentric task in the same run. As the background required to perform the allocentric task is similar in both the SSA and SLB conditions, participants may engage attention and inhibition functions in order to be able to shift between the two tasks, thereby, increasing cognitive load and thus prefrontal activity. This may explain the lack of prefrontal activity in our study, in which we used a separate background for each task.
By analyzing and comparing two tasks involving the egocentric reference frame, our study represents a milestone in the comprehension of egocentric frame mechanisms. We found significant differences in fMRI activity between the corporeal and extracorporeal egocentric tasks. Interestingly, the SLB task activated additional areas in the left hemisphere, including the angular (A39rd, A39rv, A40c) and precentral gyrus (A6cdl, A4ul). The angular gyrus is a high-level cognitive area involved in attentional and memory processes [47,48], whereas the premotor cortex is engaged in movement selection and preparation [49]. The left lateralization of premotor cortex activity during the SLB task can be viewed from the perspective of motor imagery tasks that typically activate the left premotor cortex [50][51][52]. This sheds light on the complexity of corporeal task processes. Moreover, the SLB task appears to require mental spatial coordinate transformation, as participants have to mentally align the external cues (the light bar) with their body's midsagittal plane [53]. This view is supported by studies showing the implication of left parietal temporal occipital junction lesions in body schema disturbances, causing difficulties in spatial orientation [47]. Moreover, our results are consistent with the hypothesis of a functional interhemispheric connection involved in egocentric processing, particularly during corporeal tasks [15].
The present study had several limitations. First, the sample size is small and increasing sample sizes are desired. However, previous studies have included even smaller samples (fewer than 10 participants). Moreover, the age range is wide, but we do not expect variability in healthy controls as regards this kind of task [54][55][56]. Indeed, age has been reported to affect allocentric tasks while preserving egocentric strategies. Furthermore, we did not observe any age effect on task performance. Second, we focused our investigation on the egocentric reference frame in particular. Given the lack of study on this subject, we aimed to improve knowledge of the elaboration of the egocentric reference frame. Thus, we implemented two egocentric tasks oriented toward corporal or extracorporeal spaces with a control condition when the bar changed direction. In order to verify how specific the MRI findings are in terms of brain processing of egocentricity, it would have been interesting to add an allocentric task and compare our results to previous works. Third, the 3 min runtime may have been too short to obtain reliable results. However, we used an AB block design to compensate for the short duration of the tasks, which was a requirement in this clinical setting involving patients with stroke.

Conclusions
This study investigated two egocentric tasks oriented toward either extracorporeal or corporeal spaces in 17 healthy participants. There was a large overlap between the SSA and SLB tasks, with a right fronto-parieto-occipital pattern of activity, showing that they share a common network. However, the SLB task elicited additional activity in the left angular gyrus and premotor cortex, suggesting greater complexity for SLB processes with functional interhemispheric connectivity. These results shedding light on the clinical dissociations reported for spatial cognitive disorders need to be confirmed in studies involving patients with spatial cognition disorders.
Supplementary Materials: The following are available online at https://www.mdpi.com/article/10 .3390/brainsci11080963/s1, File S1. Video of part of SSA in fMRI (gif). File S2. Video of part of SLB in fMRI (gif). Funding: This study was supported by grants from "la fondation de l'avenir" and the French Ministry of Health (PHRC AVC POSTIM/2011/09.11).
Institutional Review Board Statement: The study was conducted according to the guidelines of the Declaration of Helsinki and approved by the Poitiers III West review Board (no. 05/12/16).

Informed Consent Statement:
All participants gave their informed consent to participate in the AVCPOSTIM study.

Data Availability Statement:
The datasets used and analyzed during the current study are available from the corresponding author on request.