Access to this full-text is provided by Springer Nature.
Content available from Scientific Reports
This content is subject to copyright. Terms and conditions apply.
1
Vol.:(0123456789)
Scientic Reports | (2021) 11:23688 | https://doi.org/10.1038/s41598-021-02957-x
www.nature.com/scientificreports
The structure of the superior
and inferior parietal lobes predicts
inter‑individual suitability
for virtual reality
Chihiro Hosoda1,2*, Kyosuke Futami3, Kenchi Hosokawa2,4, Yuko Isogaya5, Tsutomu Terada6,
Kazushi Maruya4 & Kazuo Okanoya1
The global virtual reality (VR) market is signicantly expanding and being challenged with an
increased demand owing to COVID‑19. Unfortunately, VR is not useful for everyone due to large
interindividual variability existing in VR suitability. To understand the neurobiological basis of this
variability, we obtained neural structural and functional data from the participants using 3T magnetic
resonance imaging. The participants completed one of two tasks (sports training or cognitive task)
using VR, which diered in the time scale (months/minutes) and domain (motor learning/attention
task). Behavioral results showed that some participants improved their motor skills in the real world
after 1‑month training in the virtual space or obtained high scores in the 3D attention task (high
suitability for VR), whereas others did not (low suitability for VR). Brain structure analysis revealed
that the structural properties of the superior and inferior parietal lobes contain information that can
predict an individual’s suitability for VR.
Presently, the technology of immersive virtual reality (VR) is spreading exponentially in many elds of daily
life, beyond the entertainment sector. In the new “COVID-19” era, VR will be used in an increasing number of
situations. Particularly, the usefulness of VR was demonstrated in several elds as a new method for expanding
or recovering human ability, such as specialized skill acquisition (e.g., surgical training1, sports training2, educa-
tion for healthy people3, health care4, rehabilitation5, therapy for mental illness6, pain7, Parkinson’s disease8, and
children with a developmental disability9).
As the elds in which VR use for expanding or recovering human ability increase, the certainty that VR use
is an eective method for most people is crucial. Nevertheless, prior work reported that manipulation in the VR
space did not have a practical impact on all people, namely, suitability for VR use varied signicantly among
individuals. ere is marked interindividual variability in the suitability for the use of VR. In fact, several factors
have been proposed to modulate the suitability for VR use, such as gender10, preference for the VR content11,
competitive spirit12, anxiety, levels of self and spatial embodiment, sensorimotor rhythm desynchronization13,
and simulator sickness symptoms14.
However, an essential question remained unanswered in all of those empirical observations—what is the
neuronal mechanism underlying the suitability for VR? Likely, the superior parietal lobule (SPL), inferior pari-
etal lobule (IPL), and basal ganglia underly the VR suitability. Although there is no direct evidence for their
involvement in the suitability for VR, the SPL and IPL have been implicated in functions that are relevant to the
suitability for VR and that require the extraction of three-dimensional (3D) shape representations for manipu-
lating objects physically15. e SPL is involved in spatial tracking16,17 and grasping and eye movements18 and is
key to stereopsis18 and depth perception19,20.
OPEN
1Department of Life Science Graduate School of Arts and Sciences, The University of Tokyo, 3-8-1 Komaba,
meguroku, Tokyo 153-8902, Japan. 2Advanced Comprehensive Research Organization, Teikyo University,
Tokyo 173-0003, Japan. 3College of Information Science and Engineering, Ritsumeikan University, Kusatsu,
Shiga 525-8577, Japan. 4Human Information Science Laboratory, Communication Science Laboratories,
Nippon Telegraph and Telephone Corporation, Atsugi, Kanagawa 243-0198, Japan. 5Department of Psychiatry,
Tohoku Medical and Pharmaceutical University Hospital, Sendai, Miyagi 983-8512, Japan. 6Department
of Engineering, Graduate School of Engineering, Kobe University, Kobe, Hyogo 657-8501, Japan. *email:
chihirohosoda@g.ecc.u-tokyo.ac.jp
Content courtesy of Springer Nature, terms of use apply. Rights reserved
2
Vol:.(1234567890)
Scientic Reports | (2021) 11:23688 | https://doi.org/10.1038/s41598-021-02957-x
www.nature.com/scientificreports/
Other factors that support this idea include the involvement of the parietal lobe and occipital lobe in 3D visual
fatigue21, and the association between the basal ganglia, especially the caudate nucleus (CN), and dizziness22
and motion sickness23. Moreover, the CN is critically involved in memory for visual object information, such
as the encoding of spatial and response attributes (egocentric localization)24. Furthermore, patients with stroke
involving the basal ganglia, including the CN, exhibit loss of attention and spatial neglect25.
e present study aimed to address the hypotheses from a brain structural perspective, for the following
reasons. First, because some individuals have signicantly lower suitability for VR, i.e., using long-term sports
training in the virtual space, people without VR suitability cannot improve their performance in the real world,
and only those with VR suitability can expand their abilities. We hypothesized that this can be attributed to dif-
ferences in brain structure and function, especially in the SPL, the IPL, and the CN. To address this hypothesis,
we administered long-term VR sports training and tested whether patterns in specic brain areas, particularly
the SPL, IPL, and CN, predicted the ability/inability to acquire a high benet from the VR training (high/low
VR suitability).
Second, because high visuospatial attention regarding depth perception in the VR space is an essential element
for improving performance in the real world26, we applied the multiple object tracking (MOT) task involving
the frontoparietal and temporal systems27,28 with 2D and 3D conditions to examine individual dierences in
attentional functions in depth. Subsequently, to verify the generalization in another task of the predictor for VR
suitability, we tested whether the predictor of VR suitability created from VR sports training could also predict
the suitability for VR short-term attention tasks (MOT).
Results
Half of the individuals who were trained in serve–return in the virtual space exhibited improve‑
ment of the skill in the real world. To assess whether VR-based training had the eect of improving
return skills in real-world badminton players, a 4week daily VR-based serve–return training intervention was
conducted. Expectedly, not all participants in VR training exhibited improvement in their serve–return abilities
in the real world. e histogram of the improvement rate showed a binominal distribution (Fig.1a), and the
Figure1. Behavioral and imaging results of serve–return training with VR. (a) Pre–post change rate of the
return test in the actual gym aer VR serve–return training. (b) e boxplot shows the signicant dierence
in the change rate of pre–post VR training for the two groups, which were clustered according to the change
rate. (c) e high VR suitability group in the serve–return training had a signicantly greater GM volume in
the right SPL (P < 0.05, FWE-corrected). (d) More organized ber connectivity beneath the right IPL (right)
and the occipital lobe was observed in the high VR suitability vs. the low VR suitability groups (P < 0.05, FWE-
corrected). (e) e high VR suitability group in the serve–return training had a signicantly greater IPL–caudate
nucleus functional activity (P < 0.05, FWE-corrected).
Content courtesy of Springer Nature, terms of use apply. Rights reserved
3
Vol.:(0123456789)
Scientic Reports | (2021) 11:23688 | https://doi.org/10.1038/s41598-021-02957-x
www.nature.com/scientificreports/
average pre- to post-serve–return training change rate in the actual badminton court was 31.1% (SD 27.7%)
in the training group. e subjects in the control group exhibited a pre- to post-serve–return training change
rate in the actual badminton court of 5.3% (SD 7.3%). e subjects who underwent VR training were clustered
into two groups by applying the k-means method to the change rate. e number of clusters set in this study
(k = 2) was reasonable because our purpose was to classify the existence of VR suitability and its silhouette score
(more than 0.72)29. Subjects with a high growth rate were dened as the high VR suitability group; conversely,
those with low growth rate scores were dened as the low VR suitability group. Consequently, 24 individuals
had high VR suitability, and 20 had low VR suitability. e box-beard diagram depicted in Fig.1b shows the
results of these two clustered groups and control group. e mean growth rate was 2.4% in the low VR suitability
group (maximum, 10.0%; minimum, − 9.7%), 51.6% in the high VR suitability group (maximum, 90%; mini-
mum, 31.6%), and 5.3% in the control group (maximum, 10.7%; minimum, − 10.8%). Signicant dierences
between the groups (high VR suitability, low VR suitability, and control) were detected via one-way ANOVA
(F(2,55) = 114.41; P = 0.001). Tukey’s post hoc test revealed that the change rate of the high VR suitability group
(51.6% ± 3.4%) was signicantly higher than that of the low VR suitability group (2.4% ± 1.2%, P = 0.001) and
the control group (5.3% ± 1.8%, P = 0.001). No signicant dierences in visual acuity were detected among the
groups (P = 0.9).
The right IPL structure predicted VR suitability in sports training. We analyzed the participants’
brain structure using T1-weighted magnetic resonance imaging (MRI), diusion-weighted MRI, and resting-
state connectivity before the training. Dierences in brain structure across the whole brain were compared
between the high VR suitability and the low VR suitability groups. Compared with the low VR suitability, the
gray matter (GM) volume in the right SPL in high VR suitability was signicantly greater (P < 0.05, family-
wise error [FWE]-corrected) (Fig.1c, Table1). Additionally, a signicant increase in fractional anisotropy (FA)
was detected from beneath the right IPL to the occipital lobe in the high VR suitability compared to the low
VR suitability groups (P < 0.05, FWE-corrected) (Fig.1d, Table2). Moreover, signicantly increased functional
connectivity was observed between the IPL and CN in the high VR suitability compared with the low VR suit-
ability groups (Fig.1e). Because the MRI scans were obtained before the VR training, these ndings suggest that
individual dierences in the right SPL and IPL structures and in IPL-CN functional connectivity have predictive
value regarding whether an individual would have suitability for VR.
us, we evaluated whether the brain features (the GM in the right SPL, the FA in the nearby IPL, and the
IPL-CN functional connectivity) could predict the high or low VR suitability. We extracted both the GM and
the FA values from the SPL and IPL, and the resting-state values from the IPL-CN, as regional features exhibit-
ing group-wise dierences. For the dataset, the explanatory variables were these three brain features, and the
predictive variables were the two groups (high or low VR suitability). All prediction algorithms (random forest,
support vector machine, and k-NN) could predict the VR suitability at more than 80%, with the random forest
algorithm showing the highest prediction accuracy (90%, F value = 0.9; Tables3, 4). ese results indicated that
Table 1. Dierences in gray matter volume between the individuals with high VR suitability and low VR
suitability for each experiment. e coordinates (x, y, z) indicate local maxima in each brain region according
to the MNI template.
Anatomical location
Coordinates
Z-value P correctedx y z
Serve–return training
Right superior parietal lobe 20 − 49 66 4.66 0.01
Multiple object tracking
Right superior parietal lobe − 20 31 − 58 52 6.34 0.00
Table 2. Dierences in FA between the individuals high VR suitability and low VR suitability for each
experiment. e coordinates (x, y, z) indicate local maxima in each brain region according to the MNI
template.
Anatomical location
Coordinates
Z-value P correctedx y z
Serve–return training
Right superior parietal lobe 20 − 49 66 4.66 0.01
Right inferior occipital lobe 37 − 67 9 5.15 0.03
Multiple object tracking
Right superior parietal lobe 20 31 − 58 52 6.34 0.00
Right inferior occipital lobe −36 −76 9 4.45 0.03
Content courtesy of Springer Nature, terms of use apply. Rights reserved
4
Vol:.(1234567890)
Scientic Reports | (2021) 11:23688 | https://doi.org/10.1038/s41598-021-02957-x
www.nature.com/scientificreports/
the high or low suitability for VR sports training could be predicted with high probability from the structures
of the IPL, SPL and from the IPL-CN functional connectivity.
Finally, analyses of the MRI data (post vs. pre) of the high VR suitability group identied training-induced
increases in the GM of the supplemental motor area (SMA) and thalamus compared with the Control Group
(CG) (signicant time-by-group interaction). e training-induced increases in GM volume in the SMA and
thalamus were correlated with those of the pre- to post-training change rate in return score (r = 0.41; P = 0.03;
r = 0.49, P = 0.03). ese results indicated that the neural basis of the improvement in the serve–return ability in
badminton diers from that of VR suitability.
Generalization of the VR suitability predictor to 3D MOT. e results of the VR sports training
suggested that the neural correlates of the VR suitability might be the SPL, IPL, and CN, which are responsible
for the stereoscopic vision and depth perception18. To investigate the individual dierences in the information
procession of a dynamic 3D scene, we used the 2D and 3D MOT tasks. In the MOT, observers are asked to track
pre-specied multiple moving objects (targets) by visual attention among the other moving objects of identical
appearances to the target objects. e MOT involved the frontoparietal and temporal systems subserving object
recognition, attention, and working memory27,30. e subjects performed 10 trials for 2D and 3D tasks each,
and the subjects received 1 point per trial if they answered correctly; the total score was used (perfect score, 10).
Finally, we tested whether the predictor of VR suitability created from VR sports training also predicts the suit-
ability for dierent VR tasks (MOT).
A signicant dierence was detected between the 2D and 3D MOT scores (t-test, P = 0.001). On average,
the score on the 2D MOT was 8.5 (SD, 1.2; maximum, 10; minimum, 8), whereas that on the 3D MOT was 6.8
(SD, 2.0; maximum, 10; minimum, 3). e distribution was bimodal in the 3D MOT (Fig.2a), whereas it was
unimodal in the 2D MOT (Fig.2b). e subjects were clustered into two groups by applying the k-means method
to the score of the 3D MOT task (silhouette score, 0.7). Consequently, 21 individuals were in the high 3D MOT
group (male:female ratio, 12:9; average score, 0.84; maximum, 10; minimum, 7), and 17 were in the low 3D MOT
group (male:female ratio, 10:7; average score, 4.7; maximum, 6; minimum, 3). ere were signicant dierence
in the 3D MOT score in high 3D MOT group and low 3D MOT group (t-test, P = 0.01) (Fig.2c). No signicant
dierence in visual acuity was detected between the two groups (t-test, P = 0.9). When the same subjects were
clustered into two groups on the basis of the score on the 2D MOT task, no signicant dierence was detected
in the score (Fig.2d). ese results suggested that the inter-individual dierence in attentional function in depth
mightbe one of the core functions of VR suitability.
Next, we evaluated whether a predictor of suitability for long VR sports training could also predict the ability
for another short-term 3D or 2D attention VR task. e predictive model was created using data of all subjects in
VR sports training experiments, including the explanatory variables of the brain regions (GM in the SPL, FA in
the IPL, and connectivity in IPL-CN) and the predictive variables of VR suitability (high or low VR suitability)
(see “Methods”). For this purpose, we evaluated whether the VR suitability predictor from the VR sports training
could discriminate the high 3D MOT from the low 3D MOT groups. e SVM showed a prediction accuracy
of 81% (F value, 0.8; Tables5, 6). ese results suggest that VR suitability in the two tasks with dierent time
scales and contents might have a common neural basis. is VR suitability predictor failed to discriminate the
low from the high 2D MOT.
Finally, to test further the hypothesis that the common neural basis in the dierent tasks of VR suitability lies
in the IPL, SPL, and CN, a reverse verication was performed. We tested whether the predictor of VR suitability
from 3D MOT could predict VR suitability in sports training. e predictor of VR suitability in the 3D MOT
could predict the VR suitability in sports training with 82% accuracy (SVM: F value, 0.9; Tables7, 8). ese
results strongly support our hypothesis that the common neural basis in the dierent tasks of VR suitability lies
in the IPL, SPL, and CN.
Table 3. e accuracy rate of prediction for the VR suitability in the long-term serve–return training in each
algorithm.
SVM k-NN Random forest
81% 81% 90%
Table 4. e precision, recall rate, and F-values of prediction for the VR suitability in the long-term serve–
return training in each algorithm. P precision, R recall rate, F F value.
SVM k-NN Random forest
P R F P R F P R F
Low 0.85 0.77 0.81 0.85 0.77 0.81 0.90 0.90 0.91
High 0.78 0.85 0.81 0.78 0.85 0.81 0.90 0.90 0.90
Ave 0.81 0.81 0.81 0.81 0.81 0.81 0.90 0.90 0.90
Content courtesy of Springer Nature, terms of use apply. Rights reserved
5
Vol.:(0123456789)
Scientic Reports | (2021) 11:23688 | https://doi.org/10.1038/s41598-021-02957-x
www.nature.com/scientificreports/
Figure2. Behavioral results of 2D and 3D MOT. (a) e score on the 3D MOT (the number of correct answers
out of 10 trials) showed a bimodal distribution. (b) e 2D MOT score (the number of correct answers out of 10
trials) showed a normal distribution. (c) e boxplot shows the signicant dierence in the 3D MOT score for
the two groups, which were clustered according to the number of correct answers. (d) No signicant dierences
were observed in the 2D MOT score for the two groups, which were clustered according to the number of
correct answers.
Table 5. e accuracy rate of prediction for the 3D multiple object tracking from the VR suitability predictor
by the sports training in each algorithm.
SVM k-NN Random forest
81 74 76
Table 6. e precision, recall rate, and F-values of prediction for the 3D multiple object tracking from the VR
suitability predictor by the sports training in each algorithm. P precision, R recall rate, F F value.
SVM k-NN Random forest
P R F P R F P R F
Low 0.71 0.95 0.82 0.67 0.95 0.78 0.69 0.95 0.80
High 0.93 0.64 0.76 0.92 0.55 0.69 0.93 0.59 0.72
Ave 0.83 0.79 0.79 0.80 0.74 0.73 0.81 0.77 0.76
Content courtesy of Springer Nature, terms of use apply. Rights reserved
6
Vol:.(1234567890)
Scientic Reports | (2021) 11:23688 | https://doi.org/10.1038/s41598-021-02957-x
www.nature.com/scientificreports/
Serve–return training with VR induced neuroplastic changes in subjects with high VR suitabil‑
ity. To explore the training-induced changes observed in the subjects with high VR suitability in badminton
serve–return training, we conducted a 2 × 2 mixed repeated-measures ANOVA using time (before and aer
training) as a within-subject variable and group (with VR and without VR) as a between-subject variable. ese
analyses identied training-induced increases in GM of the supplemental motor cortex and thalamus (Fig.3a),
and increases in FA in the cerebellum. e training-induced increase rate in GM volume in the SMA was cor-
related with that of the serve–return ability score (r = 0.41, P = 0.03) (Fig.3b).
Discussion
We conducted a long-term VR-based badminton serve–return training among experienced badminton play-
ers and demonstrated that only subjects with high VR suitability exhibited improvement of their skills in the
real world. Moreover, we demonstrated for the rst time that individual dierences in VR suitability could be
predicted with high probability based on the brain functions and structures of the SPL, IPL, and CN. is “VR
suitability predictor” could also discriminate subjects with high performance from those with a low performance
in a VR-based 3D attention task.
A growing body of evidence indicates that the SPL and IPL are consistently active and are necessary for spatial
task performance31,32 and binocular depth19,33. e SPL is involved in computing target positions in the egocentric
reference frame for the immediate control of reaching, grasping, and eye movements31,32 and is suggested to be
one of the higher centers involved in stereopsis18. In turn, the IPL plays an important role in sustained attention18
and is engaged in action planning34,35, visuomotor control36,37, encoding38, storage39, and representation of action
sequences40. ese ndings prompted us to explore SPL and IPL structural and functional dierences between
the subjects with VR suitability and those without VR suitability.
Besides localized structural dierences in the GM volume in the SPL between individuals with high and low
VR suitability, we detected dierences in FA values extending from the occipital to the parietal lobe, including
Table 7. Prediction accuracy of VR suitability in sports training from the VR suitability predictor by
3D-multiple object trucking in each algorithm.
SVM KNN Random forests
79 82 73
Table 8. e precision, recall rate, and F-values of VR suitability in sports training from the VR suitability
predictor by 3D-multiple object trucking in each algorithm. P precision, R recall rate, F F value.
SVM KNN Random forests
P R F P R F P R F
Low 0.94 0.68 0.79 1.00 0.68 0.81 0.93 0.59 0.72
High 0.68 0.94 0.79 0.70 1.00 0.82 0.62 0.94 0.75
Ave 0.83 0.79 0.79 0.87 0.82 0.81 0.80 0.74 0.73
Figure3. GM changes aer serve–return training in the SMA within the group with high suitability for VR. (a)
Changes in GM in the SMA before and aer serve–return training in the group with high suitability for VR (b)
A correlation was found between the change rate in serve–return score and the change rate of SMA.
Content courtesy of Springer Nature, terms of use apply. Rights reserved
7
Vol.:(0123456789)
Scientic Reports | (2021) 11:23688 | https://doi.org/10.1038/s41598-021-02957-x
www.nature.com/scientificreports/
the nearby IPL. e depth perception on stereoscopy is not localized only in the SPL and IPL, as it is also asso-
ciated with early occipital areas41. Importantly, Nishida etal. showed that stereopsis leads to activation both
the occipital lobe and the SPL41. e visual system consisted of two main subsystems, originate in the primary
visual area (V1) and project ventrally to the inferotemporal cortex (ventral stream) and dorsally to the posterior
parietal cortex (dorsal stream)16. ese streams is involved in the localization of the objects in the space and
in attention. e ventral stream is engaged in the computation of the relative disparity, on the other hands, the
dorsal stream is mainly engaged in elaborating the visual input to mediate the visual control of skilled actions42
and in the computation of stereo depth based on a computation of the binocular correlation between the images
captured by the le and right eyes43. Moreover, another study showed that the early involvement of the visual
occipital cortex in the perception of motion in depth stimuli was followed by activation within the parietal cortex,
presumably associated with attention information processing44. Here the high VR suitability subjects showed
a greater GM volume in the SPL and IPL, a higher functional connectivity from the SPL to the basal ganglia,
and higher FA values along the visual pathway, suggesting that these subjects have better depth perception and
perceptual–motor coordination; thus, they showed higher suitability for VR training, especially in serve–return.
is hypothesis is also consistent with the classic ndings that damage to the posterior parietal area, including
the IPL and SPL, results in impairments in depth perception45.
Several roles of visual attention are common to MOT tasks and service–return training using VR, such as
recording the target position in a time- or space-divisional manner, maintaining and recording information
processing for the characteristics of the target, and indexing the targets46. In these models, attention may play a
role in both adding and facilitating information processing for the target and inhibiting information processing
for nontargets47. In this regard, ERP measurements during the MOT task have shown that attention was primar-
ily directed to the target48,49. Considering these studies, in the information processing in the 3D dynamic scene,
among the subjects with higher suitability for VR, the attention to the depth might be modulated to facilitate the
target processing in specic contexts, which plays an important role in the VR space. Conversely, for those who
had low suitability for VR, the depth of information might have become a distractor for facilitating the target
processing in specic contexts.
In immersive virtual environments, the distance is underestimated50,51 because technological characteristics of
HMD make presentation of precise depth cues problematic such as an accommodation-convergence mismatch.
ere were likely to be individual dierences in the inaccuracy of distance judgment in VR space50,51.
On the other hand, human vision system weights each of the depth cues, such as binocular disparity, texture
gradients, shading, motion parallax, and accommodation, and combines them to obtain a reliable 3D structure.
e degree of this weighting could be varied by feedback learning52. It is conceivable that individual dierences
in visual experience (learning) accumulated on a daily basis might produce individual dierences in the weight-
ing ratios of the various cues required to construct the 3D structure. erefore, the individual dierences in
VR suitability might relate to the individual dierence in the misperceive of the absolute egocentric distances
in current HMD systems and the cue combination in multiple depth cues (i.e. the low VR acceptability may be,
at least partly, explained by the high dependency of depth cues that are hardly available in the HMD displays).
Subjects with high VR suitability had a higher FA beneath the right IPL and the occipital lobe in the present
study. e vertical occipital fascia (VOF), which is involved in depth perception and motion perception53,54, is
a pathway that connects the V3 and the V5/MT+ in the dorsal visual pathway and the V4 in the ventral visual
pathway55,56. In the visuomotor coordination task, the VOF was related to task performance57. Furthermore,
Howells etal. found that the frontoparietal tracts, specically the dorsal branch of the superior longitudinal
fasciculus, which is responsible for visuospatial integration and motor planning, were involved in high levels of
visual–motor coordination in the right upper limb58. e present result is consistent with these ndings regard-
ing depth perception, motion perception, and visuomotor coordination, which might be an important factor
supporting VR suitability.
In the individuals with high VR suitability, the neural plasticity consisting in the increase in the GM and FA
values is correlated with the performance improvement aorded by training59–61. is result indicated that the
IPL and SPL were not involved in task-specic returning abilities; rather, they participated in VR suitability. A
previous study reported structural changes in the cerebellum aer badminton training62. Furthermore, players
of racket sports are known to perform better on visuomotor tasks, which was related to the frontal and temporal
areas63. ese results also support our hypothesis.
Although the underlying neurobiological mechanisms remain unknown, the factors proposed to explain
the changes in FA include the proportion of crossing bers, axonal permeability, and cell density or axonal/
dendritic arborization64, whereas those for GM changes include neurogenesis, gliogenesis, synaptogenesis, and
vascular changes. However, the exact biological substrates underlying the changes in GM and FA remain under
investigation, and the mechanisms underlying the increase in FA values, especially in the converters, require
future investigation.
Although two studies have conrmed the relationship between VR suitability and the structure of the IPL and
SPL, our study was limited by the fact that we were unable to assert causality between these variables, because
we did not perform a knockout experiment to prove that the disruption of IPL and SPL function increases VR
suitability. Moreover, we could not test the suitability for VR use in long-term cognitive tasks (e.g., math and
language learning) or motor tasks other than the serve–return task or whether the promotion of functional
and structural plasticity in the IPL and SPL using techniques such as the brain–machine interface will increase
VR suitability. Hence, further research is required to answer these questions. To the best of our knowledge, the
present ndings provide the rst direct evidence of the association between high suitability for VR and the IPL
and SPL. We propose that the use of the structure of the IPL and SPL as an index of VR suitability may promote
the development of more ecient VR applications for education and rehabilitation purposes in the near future,
to meet the growing demand for a wide range of VR applications.
Content courtesy of Springer Nature, terms of use apply. Rights reserved
8
Vol:.(1234567890)
Scientic Reports | (2021) 11:23688 | https://doi.org/10.1038/s41598-021-02957-x
www.nature.com/scientificreports/
Methods
Subjects. All participants were volunteers for this study and were selected from the university student (the
subjects of the serve–returning training experiment were the member of the badminton club). Aer obtaining
approval from the reviewing Board of University of Tokyo (approval no. 348-5) and Teikyo University School of
Medicine for human studies (approval no. 0326-1), our study was conducted. Standard ethical guidelines laid
down by Declaration of Helsinki-World Health Organization was followed for all methods in this study. All the
participants were informed about the study well in advance and MRI safety precautions were explained to all.
Before the study, a signed-in informed consent form was obtained from each participant.
In the serve–return training experiments, we enrolled forty three university students participating in a bad-
minton club (19 females) with a mean age of 19.8years (SD ± 3.2years; range 18–22years) as a learning group
for the serve–returning training using VR over 4weeks. To examine the eect of skill improvement, skilled
players were selected as subjects, rather than beginners. e average number of years of experience was 5.4
(SD ± 6.2; range 3–11). Fieen subjects also in a badminton club (seven females) with a mean age of 19.5years
(SD ± 0.9years; range 19–21years) participated in the study as a control group.
In the multiple objects tracking experiments, we enrolled thirty eight university students (21 males; average
age, 20.0years; SD ± 0.9years; range 18–21years) to perform the 2D and 3D MOT using VR. All subjects were
chosen through an interview and were highly motivated university students who were healthy and neurologically
intact and had no history of neuropsychiatric disorders, psychotropic medication use, or head injury.
e subjects were rigorously trained to maintain xation throughout the experiment and performed about
ve trials of the 2D and 3D MOT tasks, to ensure correct understanding and performance of the task. e order
of the 2D and 3D tasks was randomized for each subject. We performed 10 trials for each task, and the subjects
received 1 point per trial if they answered correctly; the total score was used (perfect score, 10).
Experimental design. Before the tasks (serve–return training or 2D and 3D MOT task), all participants
underwent MRI scanning (T1-weighted imaging and diusion-weighted imaging [DWI]) using a 3T MRI scan-
ner (Siemens PRISMA, Erlangen, Germany). Participants in the serve–return training also underwent MRI
scanning aer training.
Behavioral data acquisition. Serve–return training. To assess the eects of sports training using VR,
previous studies applied a training schedule spanning days or weeks2,65. In the present study, the subjects under-
went a 4week serve–return training using VR (HTC VIVE PRO; visual angle, 110°; refresh rate, 90Hz; resolu-
tion, 2880 × 1600 pixels) with racket (HTC racket handle). On each day of the training period, the participants
were required to return 80 serves to the edge of the opponent’s court in the VR space. e subject obtained 10
points if the shuttle fell within 10cm from the edge of the service line, and 5 points if the shuttle fell within other
areas on the court. Before and aer VR training, to determine if the ability of serve–returning was improved,
the subjects were required to undergo a similar rule of serve–return test in the real-world gym. For each of eight
serves, the subject returned the serves and got a score according to the position where the shuttle fell. e total
score of the 10 times of serve–return was used as the ability and compared it before and aer training.
e training environment in VR was set by Unity 2018.2.0f2, and the initial speed of serve was randomly
set from 30 to 100km/h. Two types of orbit calculation, the drive shot and the clear shot, were prepared for the
serve. In the drive shot, the eect of gravity was ignored, whereas the initial velocity was applied toward the
target point, and a constant velocity linear motion was adopted. In the case of the clear shot, the coordinates of
the midpoint between the target point and the start point were calculated, and the highest reached point was
calculated from the initial velocity (h = v0 × v0/2g). e Z coordinate (height direction) of the midpoint was set
to the highest point, and the initial velocity was given toward this point. e weight of the shuttle was set to 4g
and was a parabolic motion aected by gravity.
MOT. All participants underwent the 2D and 3D versions of the MOT task using VR (HCP VIVE PRO), with
10 repetitions each. Subjects were allowed ve practice runs each before performing the main trial, to under-
stand the rules and stabilize the score.
e order of execution of the 2D and 3D tasks was randomly selected and counterbalanced. For both tasks,
the percentage of correct responses was calculated as the number of correct responses aer 10 trials. e subjects
were allowed ve practice runs each before performing the main trial.
In the 2D MOT, the stimulus was presented in a gray square area (21 × 21cm) in the VR space. e task
proceeded with the following steps: (1) a white cross with a length of 1cm and a width of 0.1cm was presented
in the center of the square for 3s. (2) Eight white circles with a diameter of 1.0cm were presented anywhere
in the stimulus-presentation area (randomized for each trial) for 1.5s. At this time, the circles were presented
at a location where the centers of the circles were at least 2cm apart from each other so that the circles did not
overlap with each other. (3) Four of the eight white circles (randomly selected for each trial) turned green for
4s. (4) Aer returning to white from the green, all circles moved in the stimulus-presentation area in a random
direction at a speed of 6cm/s for 20s, before stopping. e circles moved in a straight line, and the direction in
which they started moving was random. e circle bounced back when it hit the wall of the stimulus-presentation
area. e circles bounced o each other when the centers of the circles approached each other up to 2cm so that
the circles did not overlap with each other. (5) Aer the white circles stopped moving, the four circles turned
green, and the numbers 1–4 were presented in the green circles in order, from le to right, for 3s. ree of the
four circles that changed to green were the rst ones to turn green (three out of the four circles were randomly
selected for each trial), and the remaining one was the rst one to turn white (one out of the four circles was
randomly selected for each trial). Subjects were asked to press a button to indicate the number of the circle that
Content courtesy of Springer Nature, terms of use apply. Rights reserved
9
Vol.:(0123456789)
Scientic Reports | (2021) 11:23688 | https://doi.org/10.1038/s41598-021-02957-x
www.nature.com/scientificreports/
was initially white but turned green at the end. (6) e correctness of the answer and the reaction time (ms; from
the presentation of the number to the answer) was recorded. e response was accepted only for 3s aer the
green circle and number were presented, aer which it was invalid. (7) Aer answering, a correct answer (the
circle that was initially white but turned green aer it stopped moving) was changed to red for 2s to provide
feedback to the subject. (8) All circles were turned o and only the gray stimulus-presentation area was presented
for 20s. Steps 1–8 were repeated 10 times.
In the 3D MOT, the stimulus was presented in a gray cubic area (21 × 21 × 21cm) in the VR space. e ow of
the 3D MOT task was basically the same as that of the 2D MOT. However, the stimulus was not a circle; rather, it
was a sphere of 1cm that was presented in the center of the spatial area. e moving direction of the sphere was
random, including depth. e procedures used were the same as those described for the 2D MOT.
Image data acquisition. e MRI data used in this research are obtained from 3T MRI scanner with a
64-channel phased-array receiver coil (Siemens PRISMA, Erlangen, Germany). High-resolution, three-dimen-
sional (3D), T1-weighted anatomical images were obtained with a magnetization-prepared rapid gradient echo
sequence, which was designed as follows: echo time (TE) = 2.98ms, repetition time (TR) = 1900ms, inversion
time = 990 ms, eld of view (FOV) = 192 × 176 mm, matrix size = 192 × 176, ip angle = 80°, and 1 mm3 iso-
tropic voxels. We also acquired whole-brain DWI as follows: TE = 62ms, TR = 6500ms, matrix size = 128 × 128,
FOV = 276 × 276mm, ip angle = 90°, 68 slices, 1 × 1 × 1 mm3 isotropic voxels. Field map images were acquired
in the same scanning space as that of the DWI (TE1 = 5.19ms; TE2 = 7.65ms).
We also acquired the multiband resting state, as follows: TE = 30ms; TR = 1500ms; slice thickness = 2mm;
voxel size = 2 × 2mm; 180 volumes; 80 slices; and ip angle = 90°. For the resting‐state functional MRI, partici-
pants were asked to see the xation points.
Quantication and statistical analysis. Image data analysis. GM-VBM. T1-weighted images were
subjected to voxel-based morphometry (VBM) analysis using the CAT toolbox (http:// dbm. neuro. uni- jena.
de/ cat. html) implemented in SPM12 (http:// www. l. ion. ucl. ac. uk/ spm). We implemented an optimized VBM
protocol for segmentation and normalization processes, using the DARTEL (Dieomorphic Anatomical Reg-
istration using Exponentiated Lie algebra) toolbox66 in SPM. e Montreal Neurological Institute (MNI) 152
standard brain template in CAT12 was adopted to do normalization of the standard space. We calculated the
total intracranial volume (TIV) for all scans. e extracted GM were smoothed using a 12mm FWHM kernel.
A 0.1 absolute masking threshold were applied to the VBM data.
For the serve–return and MOT experiments, we tested the hypothesis that particular brain regions were
reserved for suitability for VR. If this were the case, participants with particularly developed structures in the
brain would show suitability for VR. To test this hypothesis, we performed a two-sample t-test using MRI data
from the pretask condition in individuals with high and low suitability for VR in both the VR serve–return train-
ing experiment and MOT experiment. We performed a GLM analysis incorporating sex, age, TIV, and years of
badminton experience for the serve–return training experiment, and the severe stimulation sickness score as
covariates, to remove their confounding eects (P < 0.05, FWE-corrected), for the serve–return training experi-
ment and MOT experiment. In the serve–return training experiment, to identify the GM changes induced by VR
serve–return training, we conducted a 2 × 2 mixed repeated-measures analysis of variance using time (Pre and
Post) as a within-subject variable and group (training group and control group) as a between-subjects variable.
DWI-TBSS. Preprocessing and analysis of DWI data were conducted with the Oxford Center for Functional
MRI of the Brain (FMRIB) soware library (FSL 6.0.1; http:// www. fmrib. ox. ac. uk/ fsl/). Pre-processing consisted
in eddy-current correction, skull-stripping with the Brain Extraction Tool (BET), estimation of the diusion
tensor model at each voxel using the DTI t tool, generating FA. Aer the calculation of the FA map for each
participant, we implemented a voxel-wise statistical analysis of the FA data67 using TBSS.
Similar to the VBM analysis, we rst tested whether the WM structure before each task (serve–return training
and 3D MOT) could predict whether participants would have suitability for VR. To identify learning-induced
reorganization of the WM, we conducted a 2 × 2 mixed repeated-measures ANOVA using time as the within-
subject variable and group as the between-subject variable (P < 0.05, FWE-corrected), yielding FA changes specic
to the VR training program.
Resting-state connectivity_CONN. Spatially preprocessed resting‐state functional data were analyzed using
the Functional Connectivity Toolbox (CONN)68 running in MATLAB. CONN implements a component‐based
noise correction method to reduce physiological and extraneous noise, thus providing interpretative informa-
tion on correlated and anticorrelated functional brain networks69. Data were band-pass ltered (0.008–0.09Hz)
to reduce low‐frequency dri and noise eects70. e seeds are provided in the CONN soware and represent
the core and reproducibly demonstrated topological nodes within each resting-state network70. We investigated
the functional networks generated from seed (IPL). To give maps of voxel‐wise functional connectivity for each
seed ROI for each subject, the resulting coecients were converted to normally distributed scores using Fisher’s
transformation70. e value of each voxel throughout the whole brain represents the relative degree of functional
connectivity with each seed70.
Similar to the VBM and TBSS analyses, we tested whether the resting-state connectivity before each task
(serve–return training and 3DMOT) could predict whether participants would have suitability for VR. We
performed a voxel‐wise statistical analysis over the entire brain using a corrected level (P < 0.05) before a false
discovery rate correction was applied at the cluster level (P < 0.05).
Content courtesy of Springer Nature, terms of use apply. Rights reserved
10
Vol:.(1234567890)
Scientic Reports | (2021) 11:23688 | https://doi.org/10.1038/s41598-021-02957-x
www.nature.com/scientificreports/
Predictor of suitability for the VR serve–return training. We evaluated whether the brain features
(the GM in the right SPL, FA in the nearby IPL, and IPL–CN functional connectivity) could predict the level of
VR suitability. We set up a 5mm spherical VOI at the peak voxel in the right SPL for the GM, in the IPL for the
FA, and in the right CN for the resting-state activity. Next, we extracted both the GM and the FA values from
the SPL and IPL, and the resting-state values from the IPL-CN as regional features exhibiting group-wise dif-
ferences. For the dataset, the explanatory variables were these three brain features, and the predictive variables
were the two groups (high or low VR suitability). A machine learning algorithm was applied to the dataset, and
a tenfold cross-validation was performed. To validate the appropriate prediction algorithms, we used dierent
prediction algorithms, such as random forest, SVM, and k-NN. Parameters of the classier were set as follows.
Parameters of the classier parameters were set as follows so as to make the training model high accuracy. For
SVM, rbf was used as the kernel. For k-NN, the number of neighbor objects was tuned from between 1 and 5. For
random forest, gini coecients were used as a criterion for splitting. Weka (Waikato Environment for Knowl-
edge Analysis) was used for a soware for machine learning. We calculated the precision, recall, F-values for the
training group in the serve–return task and in 3D MOT.
Generalization of the predictor of VR suitability. To test the generalization of the predictor of VR
suitability, we evaluated whether the predictor of VR suitability obtained from VR sports training could dis-
criminate the high 3D MOT group from the low 3D MOT group. Firstly, we created the predictive model using
data (the GM and the FA values from the SPL and IPL, and the resting-state values from the CN) from all sub-
jects in the sports training.
Next, from the subjects who performed MOT, we extracted the GM and the FA values from the SPL and IPL,
and the resting-state values from the CN that exhibited the same coordinate of the regional features exhibiting
group-wise dierences in the VR sports training experiments. Subsequently, to test whether the VR suitability
predictor obtained from the VR sports training could discriminate the high 3D MOT from the low 3D MOT
groups, we applied the regional features from the subjects in 3D MOT to the predictor from the sports training.
To validate the appropriate algorithms, we used four dierent algorithms, i.e., random forest, SVM, and k-NN.
We also tested whether the VR suitability in 3D MOT (high 3D MOT or low 3D MOT) could be predicted
by the predictor of VR suitability created by the 3D MOT. For the dataset, the explanatory variables were these
three brain features, and the predictive variables were the two groups (high or low 3D MOT).
Finally, a reverse verication was performed to test further the hypothesis that the common neural basis in
the dierent tasks of VR suitability lies in the IPL, SPL, and CN. We tested whether the predictor of VR suitability
from 3D MOT could predict VR suitability in sports training. e predictive model was created using data from
all subjects in the sports training experiments.
Data availability
e data supporting the ndings of this study is available upon reasonable request to the corresponding author
(for verication purposes only and not for future studies). Concerning the raw brain imaging data, two partici-
pants did not agree to share, and thus data of these two participants is not available.
Received: 19 July 2021; Accepted: 24 November 2021
References
1. Frederiksen, J. G. et al. Cognitive load and performance in immersive virtual reality versus conventional vir tual reality simulation
training of laparoscopic surgery: A randomized trial. Surg. Endosc. https:// doi. org/ 10. 1007/ s00464- 019- 06887-8 (2019).
2. Kittel, A., Larkin, P., Elsworthy, N. & Spittle, M. Using 360 degrees virtual reality as a decision-making assessment tool in sport.
J. Sci. Med. Sport 22, 1049–1053. https:// doi. org/ 10. 1016/j. jsams. 2019. 03. 012 (2019).
3. Erolin, C., Reid, L. & McDougall, S. Using virtual reality to complement and enhance anatomy education. J. Vis. Commun. Med.
https:// doi. org/ 10. 1080/ 17453 054. 2019. 15976 26 (2019).
4. O’Connor, S. Virtual reality and avatars in health care. Clin. Nurs. Res. 28, 523–528. https:// doi. org/ 10. 1177/ 10547 73819 845824
(2019).
5. Antonis, J. et al. Validation of the anterior cruciate ligament (ACL) module of the VirtaMed virtual reality arthroscopy trainer.
Surg. Technol. Int. 35, 311–319 (2019).
6. Ose, S. O. et al. Exploring the potential for use of virtual reality technology in the treatment of severe mental illness among adults
in Mid-Norway: Collaborative research between clinicians and researchers. JMIR Form Res. 3, e13633. https:// doi. org/ 10. 2196/
13633 (2019).
7. Mertz, L. Virtual reality is taking the hurt out of pain. IEEE Pulse 10, 3–8. https:// doi. org/ 10. 1109/ mpuls. 2019. 29118 19 (2019).
8. Janeh, O. et al. Gait training in virtual reality: Short-term eects of dierent virtual manipulation techniques in Parkinson’s disease.
Cells https:// doi. org/ 10. 3390/ cells 80504 19 (2019).
9. Miller, I. T., Wiederhold, B. K., Miller, C. S. & Wiederhold, M. D. Virtual reality air travel training with children on the autism
spectrum: A preliminary report. Cyberpsychol. Behav. Soc. Netw. https:// doi. org/ 10. 1089/ cyber. 2019. 0093 (2019).
10. Plante, T. G., Aldridge, A., Bogden, R. & Hanelin, C. Might virtual reality promote the mood benets of exercise?. C omput. Hum.
Behav. 19, 495–509. https:// doi. org/ 10. 1016/ s0747- 5632(02) 00074-2 (2003).
11. Legrand, F. D., Joly, P. M., Bertucci, W. M., Soudain-Pineau, M. A. & Marcel, J. Interactive-virtual reality (IVR) exercise: An
examination of in-task and pre-to-post exercise aective changes. J. Appl. Sport Psychol. 23, 65–75. https:// doi. org/ 10. 1080/ 10413
200. 2010. 523754 (2011).
12. Gray, R. Transfer of training from virtual to real baseball batting. Front. Psychol. 8, 2183. https:// doi. org/ 10. 3389/ fpsyg. 2017. 02183
(2017).
13. Fowlkes, R. S. K. J. E. Simulator sickness is polygenic and polysymptomatic: Implications for research. Int. J. Aviat. Psychol. 2,
23–38 (1995).
14. Lin, J. W., Duh, H. B. L., Parker, D. E., Abi-Rached, H., Furness, T. A. In IEEE Virtual Reality Conference. 164–171.
Content courtesy of Springer Nature, terms of use apply. Rights reserved
11
Vol.:(0123456789)
Scientic Reports | (2021) 11:23688 | https://doi.org/10.1038/s41598-021-02957-x
www.nature.com/scientificreports/
15. Culham, J. C. et al. Visually guided grasping produces fMRI activation in dorsal but not ventral stream brain areas. Exp. Brain Res.
153, 180–189. https:// doi. org/ 10. 1007/ s00221- 003- 1591-5 (2003).
16. Culham, J. C. & Kanwisher, N. G. Neuroimaging of cognitive functions in human parietal cortex. Curr. Opin. Neurobiol. 11,
157–163. https:// doi. org/ 10. 1016/ s0959- 4388(00) 00191-4 (2001).
1 7. Sack, A. T. et al. e experimental combination of rTMS and fMRI reveals the functional relevance of parietal cortex for visuospatial
functions. Brain Res. Cogn. Brain Res. 13, 85–93. https:// doi. org/ 10. 1016/ s0926- 6410(01) 00087-8 (2002).
18. Sack, A. T. Parietal cortex and spatial cognition. Behav. Brain Res. 202, 153–161. https:// doi. org/ 10. 1016/j. bbr. 2009. 03. 012 (2009).
19. Ward, L. M., Morison, G., Simpson, W. A., Simmers, A. J. & Shahani, U. Using functional near infrared spectroscopy (fNIRS) to
study dynamic stereoscopic depth perception. Brain Topogr. 29, 515–523. https:// doi. org/ 10. 1007/ s10548- 016- 0476-4 (2016).
20. Breveglieri, R. et al. Functional connectivity at rest between the human medial posterior parietal cortex and the primary motor
cortex detected by paired-pulse transcranial magnetic stimulation. Brain Sci. 11(10), 1375. https:// doi. org/ 10. 3390/ brain sci11
101357 (2021).
21. Yue, K., Wang, D., Hu, H. & Fang, S. e correlation between visual fatigue and duration of viewing as assessed by brain monitor-
ing. J. Soc. Inform. Display 26, 427–437. https:// doi. org/ 10. 1002/ jsid. 667 (2018).
22. Wurthmann, S. et al. Cerebral gray matter changes in persistent postural perceptual dizziness. J. Psychosom. Res. 103, 95–101.
https:// doi. org/ 10. 1016/j. jpsyc hores. 2017. 10. 007 (2017).
23. Vignais, N., Kulpa, R., Brault, S., Presse, D. & Bideau, B. Which technology to investigate visual perception in sport: Video vs
virtual reality. Hum. Mov. Sci. 39, 12–26. https:// doi. org/ 10. 1016/j. humov. 2014. 10. 006 (2015).
24. Kesner, R. P., Bolland, B. L. & Dakis, M. Memory for spatial locations, motor responses, and objects: Triple dissociation among the
hippocampus, caudate nucleus, and extrastriate visual cortex. Exp. Brain Res. 93, 462–470. https:// doi. org/ 10. 1007/ BF002 29361
(1993).
25. Peters, F. et al. Abnormal temporal dynamics of visual attention in Alzheimer’s disease and in dementia with Lewy bodies. Neu-
robiol. Aging 33(1012), e1011–e1010. https:// doi. org/ 10. 1016/j. neuro biola ging. 2011. 10. 019 (2012).
26. Zeltzer, D. Autonomy, interaction and presence. Presence TEleoper. Virtual Environ. 1, 127–132 (1994).
27. Merkel, C., Hopf, J. M., Heinze, H. J. & Schoenfeld, M. A. Neural correlates of multiple object tracking strategies. Neuroimage 118,
63–73. https:// doi. org/ 10. 1016/j. neuro image. 2015. 06. 005 (2015).
28. Howe, P. D., Horowitz, T. S., Morocz, I. A., Wolfe, J. & Livingstone, M. S. Using fMRI to distinguish components of the multiple
object tracking task. J. Vis. 9(10), 11–11. https:// doi. org/ 10. 1167/9. 4. 10 (2009).
29. Rousseeuw, P. J. Silhouettes: A graphical aid to the interpretation and validation of cluster analysis. J. Comput. Appl. Math. 20,
53–65 (1987).
30. Nummenmaa, L., Oksama, L., Glerean, E. & Hyona, J. Cortical circuit for binding object identity and location during multiple-
object tracking. Cereb. Cortex 27, 162–172. https:// doi. org/ 10. 1093/ cercor/ bhw380 (2017).
31. Perenin, M. T. & Vighetto, A. Optic ataxia: A specic disruption in visuomotor mechanisms. I. Dierent aspects of the decit in
reaching for objects. Brain 111(Pt 3), 643–674. https:// doi. org/ 10. 1093/ brain/ 111.3. 643 (1988).
32. van der Hoorn, A., Beudel, M. & de Jong, B. M. Interruption of visually perceived forward motion in depth evokes a cortical
activation shi from spatial to intentional motor regions. Brain Res. 1358, 160–171. https:// doi. org/ 10. 1016/j. brain res. 2010. 08.
050 (2010).
33. Buckthought, A. & Mendola, J. D. A matched comparison of binocular rivalry and depth perception with fMRI. J. Vis. https:// doi.
org/ 10. 1167/ 11.6.3 (2011).
34. Jubault, T., Ody, C. & Koechlin, E. Serial organization of human behavior in the inferior parietal cortex. J. Neurosci. 27, 11028–
11036. https:// doi. org/ 10. 1523/ JNEUR OSCI. 1986- 07. 2007 (2007).
35. Hadjidimitrakis, K. et al. Common neural substrate for processing depth and direction signals for reaching in the monkey medial
posterior parietal cortex. Cereb. Cortex 24, 1645–1657. https:// doi. org/ 10. 1093/ cercor/ bht021 (2014).
36 . Buneo, C. A. & Andersen, R. A. e posterior parietal cortex: Sensorimotor interface for the planning and online control of visually
guided movements. Neuropsychologia 44, 2594–2606. https:// doi. org/ 10. 1016/j. neuro psych ologia. 2005. 10. 011 (2006).
37. Breveglieri, R., De Vitis, M., Bosco, A., Galletti, C. & Fattori, P. Interplay between grip and vision in the monkey medial parietal
lobe. Cereb. Cortex 28, 2028–2042. https:// doi. org/ 10. 1093/ cercor/ bhx109 (2018).
38. Koechlin, E., Danek, A., Burnod, Y. & Grafman, J. Medial prefrontal and subcortical mechanisms underlying the acquisition of
motor and cognitive action sequences in humans. Neuron 35, 371–381. https:// doi. org/ 10. 1016/ s0896- 6273(02) 00742-0 (2002).
39. Wu, T., Kansaku, K. & Hallett, M. How self-initiated memorized movements become automatic: A functional MRI study. J. Neu-
rophysiol. 91, 1690–1698. https:// doi. org/ 10. 1152/ jn. 01052. 2003 (2004).
40. Behrmann, M., Geng, J. J. & Shomstein, S. Parietal cortex and attention. Curr. Opin. Neurobiol. 14, 212–217. https:// doi. org/ 10.
1016/j. conb. 2004. 03. 012 (2004).
41. Nishida, Y. et al. Stereopsis-processing regions in the human parieto-occipital cortex. NeuroReport 12, 2259–2263. https:// doi. org/
10. 1097/ 00001 756- 20010 7200- 00043 (2001).
42. Milner, A. D. & Go odale, M. A. Two visual systems re-viewed. Neuropsychologia 46, 774–785. htt ps:// do i . org/ 10. 1016/j. neuro psych
ologia. 2007. 10. 005 (2008).
43. Minini, L., Parker, A. J. & Bridge, H. Neural modulation by binocular disparity greatest in human dorsal visual stream. J. Neuro-
physiol. 104, 169–178. https:// doi. org/ 10. 1152/ jn. 00790. 2009 (2010).
44. Lamberty, K., Gobbele, R., Schoth, F., Buchner, H. & Waberski, T. D. e temporal pattern of motion in depth perception derived
from ERPs in humans. Neurosci. Lett. 439, 198–202. https:// doi. org/ 10. 1016/j. neulet. 2008. 04. 101 (2008).
45. Lehmann, D. & Walchli, P. Depth perception and location of brain lesions. J. Neurol. 209, 157–164. https:// doi. org/ 10. 1007/ BF003
12538 (1975).
46. Pylyshyn, Z. W. & Storm, R. Tracking multiple independent targets: Evidence for a parallel tracking mechanism. Spat. Vis. 3,
179–197. https:// doi. org/ 10. 1163/ 15685 6888x 00122 (1988).
47. Pylyshyn, Z. W. & Annan, V. Jr. Dynamics of target selection in multiple object tracking (MOT). Spat. Vis. 19, 485–504. https://
doi. org/ 10. 1163/ 15685 68067 79194 017 (2006).
48. Drew, T., McCollough, A. W., Horowitz, T. S. & Vogel, E. K. Attentional enhancement during multiple-object tracking. Psychon.
Bull. Rev. 16, 411–417. https:// doi. org/ 10. 3758/ PBR. 16.2. 411 (2009).
49. Doran, M. M. & Homan, J. E. e role of visual attention in multiple object tracking: Evidence from ERPs. Atten. Percept. Psy-
chophys. 72, 33–52. https:// doi. org/ 10. 3758/ APP. 72.1. 33 (2010).
50. Interrante, V., Ries, B. & Anderson, L. In IEEE Virtual Reality Conference 3–10 (2006).
51. Peer, A. & Ponto, K. In IEEE Symposium on 3d User Interfaces 83–86 (2017).
52. Ernst, M. O., Banks, M. S. & Bultho, H. H. Touch can change visual slant perception. Nat. Neurosci. 3, 69–73. https:// doi. org/ 10.
1038/ 71140 (2000).
53. Yeatman, J. D. et al. e vertical occipital fasciculus: A century of controversy resolved by invivo measurements. Proc. Natl. Acad.
Sci. U. S. A. 111, E5214-5223. https:// doi. org/ 10. 1073/ pnas. 14185 03111 (2014).
54. Oishi, H., Takemura, H., Aoki, S. C., Fujita, I. & Amano, K. Microstructural properties of the vertical occipital fasciculus explain
the variability in human stereoacuity. Proc. Natl. Acad. Sci. U. S. A. 115, 12289–12294. https:// doi. org/ 10. 1073/ pnas. 18047 41115
(2018).
Content courtesy of Springer Nature, terms of use apply. Rights reserved
12
Vol:.(1234567890)
Scientic Reports | (2021) 11:23688 | https://doi.org/10.1038/s41598-021-02957-x
www.nature.com/scientificreports/
55. Yeatman, J. D., Rauschecker, A. M. & Wandell, B. A. Anatomy of the visual word form area: Adjacent cortical circuits and long-
range white matter connections. Brain Lang. 125, 146–155. https:// doi. org/ 10. 1016/j. bandl. 2012. 04. 010 (2013).
56. Takemura, H. et al. A major human white matter pathway between dorsal and ventral visual cortex. Cereb. Cortex 26, 2205–2214.
https:// doi. org/ 10. 1093/ cercor/ bhv064 (2016).
57. Budisavljevic, S., Dell’Acqua, F. & Castiello, U. Cross-talk connections underlying dorsal and ventral stream integration during
hand actions. Cortex 103, 224–239. https:// doi. org/ 10. 1016/j. cortex. 2018. 02. 016 (2018).
58. Howells, H. et al. Frontoparietal tracts linked to lateralized hand preference and manual specialization. Cereb. Cortex 28, 2482–2494.
https:// doi. org/ 10. 1093/ cercor/ bhy040 (2018).
59. Hosoda, C., Tanaka, K., Nariai, T., Honda, M. & Hanakawa, T. Dynamic neural network reorganization associated with second
language vocabulary acquisition: A multimodal imaging study. J. Neurosci. 33, 13663–13672. h t tps:// doi. org/ 10. 1523/ JNEUR OSCI.
0410- 13. 2013 (2013).
60. Hosoda, C. et al. Plastic frontal pole cortex structure related to individual p ersistence for goal achievement. Commun. Biol. 3, 194.
https:// doi. org/ 10. 1038/ s42003- 020- 0930-4 (2020).
61. Yotsumoto, Y. et al. White matter in the older brain is more plastic than in the younger brain. Nat. Commun. 5, 5504. https:// doi.
org/ 10. 1038/ ncomm s6504 (2014).
62. Shao, M. et al. Learning to play badminton altered resting-state activity and functional connectivity of the cerebellar sub-regions
in adults. PLoSOne 14, e0223234. https:// doi. org/ 10. 1371/ journ al. pone. 02232 34 (2019).
63. Guo, Z., Li, A. & Yu, L. “Neural Eciency” of athletes’ brain during visuo-spatial task: An fMRI study on table tennis players.
Front. Behav. Neurosci. 11, 72. https:// doi. org/ 10. 3389/ fnbeh. 2017. 00072 (2017).
64. Taubert, M. et al. Dynamic properties of human brain structure: Learning-related changes in cortical areas and associated ber
connections. J. Neurosci. 30, 11670–11677. https:// doi. org/ 10. 1523/ JNEUR OSCI. 2567- 10. 2010 (2010).
65. Romeas, T., Chaumillon, R., Labbe, D. & Faubert, J. Combining 3D-MOT with sport decision-making for perceptual-cognitive
training in virtual reality. Percept. Mot. Skills https:// doi. org/ 10. 1177/ 00315 12519 860286 (2019).
66. Ashburner, J. A fast dieomorphic image registration algorithm. Neuroimage 38, 95–113. https:// doi. org/ 10. 1016/j. neuro image.
2007. 07. 007 (2007).
67. Behrens, T. E., Berg, H. J., Jbabdi, S., Rushworth, M. F. & Woolrich, M. W. Probabilistic diusion tractography with multiple bre
orientations: What can we gain?. Neuroimage 34, 144–155. https:// doi. org/ 10. 1016/j. neuro image. 2006. 09. 018 (2007).
68. Whiteld-Gabrieli, S. & Nieto-Castanon, A. Conn: A functional connectivity toolbox for correlated and anticorrelated brain
networks. Brain Connect. 2, 125–141. https:// doi. org/ 10. 1089/ brain. 2012. 0073 (2012).
69. Behzadi, Y., Restom, K., Liau, J. & Liu, T. T. A component based noise correction method (CompCor) for BOLD and perfusion
based fMRI. Neuroimage 37, 90–101. https:// doi. org/ 10. 1016/j. neuro image. 2007. 04. 042 (2007).
70. Whiteld-Gabrieli, S. et al. Associations and dissociations between default and self-reference networks in the human brain. Neu-
roimage 55, 225–232. https:// doi. org/ 10. 1016/j. neuro image. 2010. 11. 048 (2011).
Acknowledgements
is study was supported by grants from PRESTO, CREST (JPMJCR18A3), and Souhatsu from Japan Science
and Technology Agency (JST) to CH.
Author contributions
C.H. conceptualized and C.H. and K.M. contributed to the design of the work. C.H. and K.F. took part in the
data acquisition. C.H. and K.F. contributed to the analysis and C.H., K.F., K.M., and K.H. interpreted the data of
the work. C.H. and F.K. contributed to draing this paper. All authors approved the nal version for submission
and agreed to be accountable for all aspects of the work to ensure that all questions related to the accuracy or
integrity of any part of the work are appropriately investigated and resolved.
Competing interests
e authors declare no competing interests.
Additional information
Correspondence and requests for materials should be addressed to C.H.
Reprints and permissions information is available at www.nature.com/reprints.
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and
institutional aliations.
Open Access is article is licensed under a Creative Commons Attribution 4.0 International
License, which permits use, sharing, adaptation, distribution and reproduction in any medium or
format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the
Creative Commons licence, and indicate if changes were made. e images or other third party material in this
article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the
material. If material is not included in the article’s Creative Commons licence and your intended use is not
permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from
the copyright holder. To view a copy of this licence, visit http:// creat iveco mmons. org/ licen ses/ by/4. 0/.
© e Author(s) 2021
Content courtesy of Springer Nature, terms of use apply. Rights reserved
1.
2.
3.
4.
5.
6.
Terms and Conditions
Springer Nature journal content, brought to you courtesy of Springer Nature Customer Service Center GmbH (“Springer Nature”).
Springer Nature supports a reasonable amount of sharing of research papers by authors, subscribers and authorised users (“Users”), for small-
scale personal, non-commercial use provided that all copyright, trade and service marks and other proprietary notices are maintained. By
accessing, sharing, receiving or otherwise using the Springer Nature journal content you agree to these terms of use (“Terms”). For these
purposes, Springer Nature considers academic use (by researchers and students) to be non-commercial.
These Terms are supplementary and will apply in addition to any applicable website terms and conditions, a relevant site licence or a personal
subscription. These Terms will prevail over any conflict or ambiguity with regards to the relevant terms, a site licence or a personal subscription
(to the extent of the conflict or ambiguity only). For Creative Commons-licensed articles, the terms of the Creative Commons license used will
apply.
We collect and use personal data to provide access to the Springer Nature journal content. We may also use these personal data internally within
ResearchGate and Springer Nature and as agreed share it, in an anonymised way, for purposes of tracking, analysis and reporting. We will not
otherwise disclose your personal data outside the ResearchGate or the Springer Nature group of companies unless we have your permission as
detailed in the Privacy Policy.
While Users may use the Springer Nature journal content for small scale, personal non-commercial use, it is important to note that Users may
not:
use such content for the purpose of providing other users with access on a regular or large scale basis or as a means to circumvent access
control;
use such content where to do so would be considered a criminal or statutory offence in any jurisdiction, or gives rise to civil liability, or is
otherwise unlawful;
falsely or misleadingly imply or suggest endorsement, approval , sponsorship, or association unless explicitly agreed to by Springer Nature in
writing;
use bots or other automated methods to access the content or redirect messages
override any security feature or exclusionary protocol; or
share the content in order to create substitute for Springer Nature products or services or a systematic database of Springer Nature journal
content.
In line with the restriction against commercial use, Springer Nature does not permit the creation of a product or service that creates revenue,
royalties, rent or income from our content or its inclusion as part of a paid for service or for other commercial gain. Springer Nature journal
content cannot be used for inter-library loans and librarians may not upload Springer Nature journal content on a large scale into their, or any
other, institutional repository.
These terms of use are reviewed regularly and may be amended at any time. Springer Nature is not obligated to publish any information or
content on this website and may remove it or features or functionality at our sole discretion, at any time with or without notice. Springer Nature
may revoke this licence to you at any time and remove access to any copies of the Springer Nature journal content which have been saved.
To the fullest extent permitted by law, Springer Nature makes no warranties, representations or guarantees to Users, either express or implied
with respect to the Springer nature journal content and all parties disclaim and waive any implied warranties or warranties imposed by law,
including merchantability or fitness for any particular purpose.
Please note that these rights do not automatically extend to content, data or other material published by Springer Nature that may be licensed
from third parties.
If you would like to use or distribute our Springer Nature journal content to a wider audience or on a regular basis or in any other manner not
expressly permitted by these Terms, please contact Springer Nature at
onlineservice@springernature.com
Available via license: CC BY 4.0
Content may be subject to copyright.
