Neural encoding of saltatory pneumotactile velocity in human glabrous hand

Abstract

Neurons in the somatosensory cortex are exquisitely sensitive to mechanical stimulation of the skin surface. The location, velocity, direction, and adaptation of tactile stimuli on the skin’s surface are discriminable features of somatosensory processing, however the representation and processing of dynamic tactile arrays in the human somatosensory cortex are poorly understood. The principal aim of this study was to map the relation between dynamic saltatory pneumatic stimuli at discrete traverse velocities on the glabrous hand and the resultant pattern of evoked BOLD response in the human brain. Moreover, we hypothesized that the hand representation in contralateral Brodmann Area (BA) 3b would show a significant dependence on stimulus velocity. Saltatory pneumatic pulses (60 ms duration, 9.5 ms rise/fall) were repetitively sequenced through a 7-channel TAC-Cell array at traverse velocities of 5, 25, and 65 cm/s on the glabrous hand initiated at the tips of D2 (index finger) and D3 (middle finger) and sequenced towards the D1 (thumb). The resulting hemodynamic response was sampled during 3 functional MRI scans (BOLD) in 20 neurotypical right-handed adults at 3T. Results from each subject were inserted to the one-way ANOVA within-subjects and one sample t-test to evaluate the group main effect of all three velocities stimuli and each of three different velocities, respectively. The stimulus evoked BOLD response revealed a dynamic representation of saltatory pneumotactile stimulus velocity in a network consisting of the contralateral primary hand somatosensory cortex (BA3b), associated primary motor cortex (BA4), posterior insula, and ipsilateral deep cerebellum. The spatial extent of this network was greatest at the 5 and 25 cm/s pneumotactile stimulus velocities.

Full text

RESEARCH ARTICLE
Neural encoding of saltatory pneumotactile
velocity in human glabrous hand
Hyuntaek Oh
1,3¤
*, Rebecca Custead
2,3
, Yingying Wang
2,3
, Steven Barlow
1,2,3
1Department of Biological Systems Engineering, University of Nebraska, Lincoln, Nebraska, UnitedStates of
America, 2Department of Special Education and Communication Disorders, University of Nebraska, Lincoln,
Nebraska, United States of America, 3Center for Brain, Biology and Behavior, University of Nebraska,
Lincoln, Nebraska, United States of America
¤Current address: Department of Neuroscience, Baylor College of Medicine, Houston, Texas, UnitedStates
of America
*hyuntaek.oh@huskers.unl.edu
Abstract
Neurons in the somatosensory cortex are exquisitely sensitive to mechanical stimulation of
the skin surface. The location, velocity, direction, and adaptation of tactile stimuli on the
skin’s surface are discriminable features of somatosensory processing, however the repre-
sentation and processing of dynamic tactile arrays in the human somatosensory cortex are
poorly understood. The principal aim of this study was to map the relation between dynamic
saltatory pneumatic stimuli at discrete traverse velocities on the glabrous hand and the
resultant pattern of evoked BOLD response in the human brain. Moreover, we hypothesized
that the hand representation in contralateral Brodmann Area (BA) 3b would show a signifi-
cant dependence on stimulus velocity. Saltatory pneumatic pulses (60 ms duration, 9.5 ms
rise/fall) were repetitively sequenced through a 7-channel TAC-Cell array at traverse veloci-
ties of 5, 25, and 65 cm/s on the glabrous hand initiated at the tips of D2 (index finger) and
D3 (middle finger) and sequenced towards the D1 (thumb). The resulting hemodynamic
response was sampled during 3 functional MRI scans (BOLD) in 20 neurotypical right-
handed adults at 3T. Results from each subject were inserted to the one-way ANOVA
within-subjects and one sample t-test to evaluate the group main effect of all three velocities
stimuli and each of three different velocities, respectively. The stimulus evoked BOLD
response revealed a dynamic representation of saltatory pneumotactile stimulus velocity in
a network consisting of the contralateral primary hand somatosensory cortex (BA3b), asso-
ciated primary motor cortex (BA4), posterior insula, and ipsilateral deep cerebellum. The
spatial extent of this network was greatest at the 5 and 25 cm/s pneumotactile stimulus
velocities.
Introduction
Animal and human models of brain plasticity have shown that the development of functional
motor tasks depend on the interplay between sensory input and motor output [1,2]. Among
PLOS ONE | https://doi.org/10.1371/journal.pone.0183532 August 25, 2017 1 / 17
a1111111111
a1111111111
a1111111111
a1111111111
a1111111111
OPEN ACCESS
Citation: Oh H, Custead R, Wang Y, Barlow S
(2017) Neural encoding of saltatory pneumotactile
velocity in human glabrous hand. PLoS ONE 12(8):
e0183532. https://doi.org/10.1371/journal.
pone.0183532
Editor: Christos Papadelis, Boston Children’s
Hospital / Harvard Medical School, UNITED
STATES
Received: April 24, 2017
Accepted: August 5, 2017
Published: August 25, 2017
Copyright: ©2017 Oh et al. This is an open access
article distributed under the terms of the Creative
Commons Attribution License, which permits
unrestricted use, distribution, and reproduction in
any medium, provided the original author and
source are credited.
Data Availability Statement: fMRI data have been
uploaded to Figshare. All interested researchers
can download the data from the website (https://
figshare.com/articles/Neural_encoding_of_
saltatory_pneumotactile_velocity_in_human_
glabrous_hand/5276992; DOI: https://doi.org/10.
6084/m9.figshare.5276992.v1).
Funding: This work was supported by the internal
grant from Barkley Trust Foundation (SM Barlow –
PI). There was no additional external funding
received for this study.

the many functions of the somatosensory system, processing information about the location,
velocity and traverse length of tactile stimuli on the body surface is presumed essential for the
development and maintenance of fine motor control of the hand [3–5]. Improving our knowl-
edge of velocity and directional encoding in this sensory domain will help formulate innova-
tive neurotherapeutic strategies for the rehabilitation of brain-damaged patients to regain
motor skills in the limb (hand, foot) and orofacial (speech, gesture, swallowing) systems [6].
Limited data exist on the cortical representation of moving touch stimulation on the glabrous
skin of the digits in humans [7,8], and many studies involving sensorimotor tasks have been
limited to neurotypical adults using electrical and/or transcranial magnetic stimulation (TMS)
[9–11].
The sensory flow of tactile information derived from mechanoreceptors in the glabrous
skin of the hand is conveyed along the dorsal column-medial lemniscus and transmitted
through the contralateral ventroposterolateral (VPL) thalamus and primary somatosensory
cortex (S1), whereas the secondary somatosensory cortex (S2) typically shows a bilateral
response to a unilateral somatosensory stimulus [12,13]. Many neurons in the posterior parie-
tal cortex (PPC) respond to both tactile and visual inputs [14,15], with select sensorimotor
transformation and output to the premotor cortex (PMC) [16]. The cerebellum represents the
‘forward model’ of the sensorimotor system that implements predictions of the sensory result
from the motor commands, and these predictions can be used to improve a motor skill or acti-
vate sensorimotor plasticity [17,18]. Several neuroimaging studies using functional Magnetic
Resonance Imaging (fMRI) and positron emission tomography (PET) have discovered that the
cerebellum is involved in signaling the sensory consequence of movements resulting from the
correlation between the actual and predicted sensory feedback, and forward models stored in
the cerebellum are related to predictions of movements [19,20]. Since the cerebellum plays an
important role in predictive motor control and storing forward models [21,22], recent human
studies highlight the crucial role of the cerebellum and sensorimotor cortex during motor
learning and functional recovery from stroke [23,24].
Moving tactile stimulation on glabrous skin, known historically as ‘surface parallel stimula-
tion’ [4], has been shown to evoke activity among cortical and subcortical somatosensory rep-
resentations [25]. Human psychophysical studies have shown that the optimal range of
stimulus velocity for the discrimination of skin traverse velocity lies between 3 and 25 cm/s
[3,4,26,27]. Similar velocities of brush stroke stimuli have been used to map the Blood-oxy-
gen-level dependent (BOLD) responses in S1 and posterior insular cortex [28,29]. Beyond this
optimal range, neurotypical subjects were still able to recognize the brushing stimuli at veloci-
ties exceeding 50 cm/s, however the perception of discrimination of velocity became less reli-
able due to changes in perceived stimulation location, direction, and distance. At low velocities
(e.g., <3 cm/s), S1 neurons appear to encode the moving tactile stimulation as discrete stimu-
lus events rather than a progressive traverse motion track. Furthermore, an accurate discrimi-
nation of skin velocity on glabrous skin of the hand may yield better encoding over a wider
range of velocities compared to the hairy skin since Aβmechanoreceptors in the glabrous skin
are superior at encoding the temporal and spatial properties of incoming stimuli [27,30,31].
Thus, a consideration of the optimal operating range for velocity and direction of moving tac-
tile stimulation on the glabrous hand are important factors to consider when designing a per-
ceptual or functional imaging experiment with human subjects [32,33].
The glabrous hand and orofacial skin feature high innervation densities, large number of
receptive fields, and acute sensitivity which translate to high cortical magnification in S1 [34].
Many neuroimaging modalities such as 1.5 T fMRI, magnetoencephalography (MEG), or PET
of the human brain do not provide enough spatial resolution to map individual fingers and
their phalanges because the distances between individual digits and segments represented in
Neural encoding of pneumotactile velocity
PLOS ONE | https://doi.org/10.1371/journal.pone.0183532 August 25, 2017 2 / 17
Competing interests: The authors have declared
that no competing interests exist.

S1 are only a few mm [35]. Thus, high resolution 3T fMRI with multichannel head coils are
better equipped to achieve small voxel size combined with precisely controlled dynamic spatial
tactile arrays to map the hand-finger somatotopy under conditions where velocity and/or
direction are independent variables of interest [36,37]. There are inherent challenges in the
design of an MRI-compatible tactile stimulus array control system that is scalable for velocity
and direction. A limited number of studies have explored tactile encoding using continuous
moving brush, piezo-element vibration, and compressed air [38–42]. Thus, in order to
advance our understanding of tactile velocity encoding networks in the human brain, the need
exists for a programmable, multichannel tactile stimulus control system that is non-invasive,
simple to configure and can be applied anywhere on the body with scalable velocity control
and fully MRI compatibility.
The primary goal of the present study was to functionally map the human brain to identify
the relation between saltatory pneumotactile stimulation at 3 velocities on the glabrous hand
and the evoked hemodynamic BOLD response in select regions of interest (ROIs), including
cerebral somatosensory areas (S1, S2, PPC, posterior insula), and deep cerebellum among 20
neurotypical adults using high-resolution fMRI methods. In this study, three velocities, includ-
ing a relatively low but not discrete velocity (5cm/s), a medium velocity (25cm/s), and a rela-
tively high but perceptable velocity (65cm/s), were chosen to investigate the dynamic BOLD
response between low and high end of perceivable velocity range. We hypothesized that the
somatosensory network would show evidence of modulation as reflected in %BOLD change
among the ROIs of interest as a function of saltatory velocity. To achieve this objective, a
7-channel TAC-Cell array developed in our laboratory was configured to the glabrous hand on
three digits, including D1 (thumb), D2 (index finger), and D3 (middle finger) for saltatory
pneumatic stimulation randomized at 3 velocities.
Materials and methods
Subjects
Twenty right-handed, neurotypical adults (14 females, 6 males) age 18-30 years (mean = 22.3
±2.47 years) participated in this study. Exclusion criteria: traumatic injury to the hand or neu-
rological disease resulting in sensorimotor impairment affecting hand movement and/or sen-
sory function. Each subject provided informed written consent in accordance with the
University of Nebraska—Lincoln institutional review board approval.
Stimulus device: Galileo somatosensory pneumatic stimulus control
system
A multichannel pneumatic amplifier and tactile array known as the Galileo Somatosensory™
system (Epic Medical Concepts & Innovations, Inc., Mission, Kansas USA) was used for
mechanosensory stimulus generation. The Galileo features scalable pulse generation in config-
urable arrays, and is fully MRI/MEG compatible. The pneumatic stimulator probes, known as
TAC-Cells, are made from acetyl thermoplastic homopolymer, use tiny volumes of com-
pressed air to rapidly deform (10 ms rise/fall times) the surface of the skin. The individual
pressure pulses generated by the Galileo controller are transmitted through 18’ of polyurethane
tubing (3/32” ID) which is routed through a waveguide into the MRI suite and terminated
with TAC-Cells to allow for placement on the subject’s hand with the bore of the MRI scanner.
The PC laptop computer, Galileo SomatosensoryTM pneumatic controller, and integrated
dual-cylinder pump motor are all located outside the shielded MRI scanner suite room. As
shown in Fig 1, the TAC-Cell is essentially a small capsule with a sealing flange (6 mm ID,
Neural encoding of pneumotactile velocity
PLOS ONE | https://doi.org/10.1371/journal.pone.0183532 August 25, 2017 3 / 17

15 mm OD), which can be adhered rapidly to virtually any skin surface, including the glabrous
hand and face [6,43,44].
fMRI data acquisition
A brain structural MRI scan and 3 functional sessions (BOLD) were recorded at 3.0 T (Skyra,
Siemens Medical Solutions, Erlangen, Germany) using a 32-channel head coil. Structural
T1-weighted 3-dimensional image of the subject’s brain (MPRAGE, Magnetization-Prepared
Rapid Gradient-Echo) was acquired at the beginning of the session [repetition time (TR) =
2400 ms, echo time (TE) = 3.37 ms, voxel size = 1 x 1 x 1 mm, flip angle = 7˚, number of
slices = 192, acquisition matrix = 256 x 256, field of view (FoV) = 256 x 256 mm, total acquisi-
tion time (TA) = 5:35 minutes].
Following the MPRAGE anatomical scan, three sessions of functional images were recorded
using a T2-weighted EPI (Echo Planar Imaging) sequence [TR = 2500 ms, TE = 30 ms, voxel
size = 2.5 x 2.5 x 2.5 mm, flip angle = 83˚, number of slices = 320, acquisition matrix = 88 x 88,
FoV = 220 x 220 mm, Phase partial Fourier factor = 7/8, TA = 13:53].
Visual countdown presentation to maintain the subjects’ vigilance was performed using E-
prime 2.0 software (Psychology Software Tools, Inc., Sharpsburg, PA, USA). This visual pre-
sentation was projected onto a screen behind (headward) the scanner bore. The subject
observed the presentation on a mirror which was attached to the 32-channel head coil. The
visual countdown presentation included a declining sequence of numbers (20:1) which
Fig 1. Galileo somatosensory™tactile stimulation. Top Left: Galileo somatosensory™tactile stimulator.
Top Right: Stimulated areas and TAC-Cells location. p1 in D2 and D3 = red, p2 in D2 and D3 = orange,
p4 in D2 and D3 = yellow, p4 in D1 = green, p1 in D1 = blue. The TAC-Cell is essentially a small capsule
(OD = 15 mm, ID = 6 mm), and machined from acetal thermoplastic. Bottom: Stimulus velocity pressure
waveforms for each condition.
https://doi.org/10.1371/journal.pone.0183532.g001
Neural encoding of pneumotactile velocity
PLOS ONE | https://doi.org/10.1371/journal.pone.0183532 August 25, 2017 4 / 17

corresponds to the number of remaining stimulus blocks in the BOLD session. The number
on the presentation was shown only for 0.5 second to minimize a primary visual cortex
response.
Tactile stimulus control
Seven small plastic pneumatic TAC-Cells (6mm ID) were placed on the palm of the right hand
along the length of index and middle finger using tincture of Benzoin (10% concentration to
increase adhesion) followed by the application of double adhesive tape collars. A Galileo™
Somatosensory tactile array was programmed to deliver punctate (60 ms duration, 9 ms rise/
fall) pneumotactile sequence through TAC-Cells placed on the glabrous skin of the right hand
(see Fig 1), including p1, p2 segments of D3 (middle finger), p1, p2, p4 segments of D2 (index
finger), and p4, p1 of D1 (thumb). Morphometric dimensions between p1 and p2 in D2
(Length 1), p2 and p3 in D2 (Length 2), p4 in D2 and P4 in D1 (Length 3), and p4 and p1 in
D1 (Length 4) were measured from each subject to adjust for variations in hand size to create
accurate tactile traverse velocities (Fig 1). Programmed time delays between individual TAC-
Cells result in a saltatory velocity sequence traversing the tips of D1, D2 through the basal pha-
langeal segments to the distal phalanx of the thumb. The silicon tubing was bifurcated at its
terminal for channels 1 and 2 to deliver a pneumotactile stimulus on the p1 and p2 segments
of the D2 and D3. Rice-filled hand-warmers placed within mitten gloves were fit to all subject’s
right hand to maintain normothermia of limb extremities during testing in the MRI scanner
suite [45]. It is through this array of pneumatically charged TAC-Cells that the subject experi-
enced repeated trains of saltatory pulsed pneumotactile stimulation ranging from very slow
(5 cm/s) to fast (65 cm/s) traverse speeds on the glabrous surface of the hand.
A randomized-balanced block design (40 sec duration/block) included the following 5 con-
ditions: Saltatory velocities @ 5, 25, and 65 cm/sec, simultaneous TAC-Cells ON, and all cells
OFF (Fig 2). There were three sessions during the fMRI BOLD response acquisition and each
session included 4 cycles of the 5 stimulus conditions. Thus, a total of 20 conditions in each
session were counter-balanced and randomized. The duration of the stimulus event for each
condition was 20 seconds (8 volumes, TR = 2500 ms), followed by 20 seconds of rest. The tac-
tile stimulus was continuously delivered from p1 of D2 and D3 to p1 of D1, passing through
both p2 of D2 and D3, p4 of D2 and D3, p4 of D2 and p4 of D1. The average time elapsed
between velocity trains of 5, 25 and 65 cm/sec were 501.1 ms, 153 ms, and 37.6 ms,
Fig 2. Experiment stimulus design. Random-balanced experimental block design in one functional scan
session. One session includes 4 cycles of the 5 stimulus conditions and the total measurement time of one
session was 13:20 min.
https://doi.org/10.1371/journal.pone.0183532.g002
Neural encoding of pneumotactile velocity
PLOS ONE | https://doi.org/10.1371/journal.pone.0183532 August 25, 2017 5 / 17

respectively. Total BOLD sampling time of one session was 13:20 min (320 volumes), thus 3
BOLD acquisitions produced 960 volumes of fMRI data per subject.
fMRI data analysis
Pre-processing and statistical analysis of MPRAGE and functional images were performed
using SPM 12 (Statistical Parametric Mapping; Wellcome Department of Imaging Neurosci-
ence, London, UK). The 3 sessions of functional MRI volumes were realigned to the first vol-
ume in the each session, normalized to adjust overall size and orientation of the functional and
anatomical images to the MNI template, and smoothed by convolution with an isotropic
Gaussian kernel (FWHM = 8 mm).
The General Linear Model (GLM) was applied to estimate the predictor variables by
convolving the design matrix (the box-car stimulus blocks) and the hemodynamic response
function for the single-subject analysis of BOLD responses from different velocities of tactile
stimulus [46]. The model includes five regressors (5 cm/s, 25 cm/s, 65 cm/s, all TAC-Cells OFF,
and all TAC-Cells ON), and six motion parameter correction regressors (three translational
axes [X, Y, Z] and three rotations [roll, yaw, pitch]) per session. One-sided main effect for each
velocity condition was determined by subtracting the no stimulus block contrast (control
block). Resulting t-maps from each BOLD session were carried forward to the Mixed Effects
(MFX) analysis to combine the 3 BOLD results within a subject [47]. An F-contrast was
required to determine the main effect of velocity conditions. The result from F-contrasts
showed how the different stimulus velocities change brain response and where the stimulus
influences the BOLD response in the brain. Besides the result from F-contrasts, six additional
contrasts were created: 1) 5 cm/s >No stimulus, 2) 25 cm/s >No stimulus, 3) 65 cm/s >No
stimulus, 4) 5 cm/s >All TAC-Cells ON, 5) 25 cm/s >All TAC-Cells ON, and 6) 65 cm/s >All
TAC-Cells ON. The contrast results from each subject were entered into the 2nd-level analysis
to access the group analysis. On SPM group analysis, an uncorrected p-value = .0001 was used.
The group analysis of one-sided main effects for 5 cm/s, 25 cm/s and 65 cm/s accepted the one
sample t-test which was used to compute within-subject contrast results from 1st-level analysis.
One-way ANOVA analysis was implemented to derive the group main effect among the various
velocity stimulus profiles. The t-contrast results from each subject were used in the one-way
ANOVA analysis.
Results
The seven TAC-Cells, configured to digits D1, D2, and D3 of the glabrous right hand which
were programmed to produce 3 saltatory velocities (5, 25, 65 cm/s) were highly effective in evok-
ing a scalable BOLD response among several ROIs within the human somatosensory network.
The first-level result from each single subject was acquired by combining 3 BOLD sessions
with the exception of one subject who had 2 BOLD sessions. The significant level was set to
P
unc
<.0001 for the five stimulus conditions (5 cm/s, 25 cm/s, 65 cm/s, All-Off (No stimulus),
and All-ON). A dominant contralateral response among the velocity conditions was consis-
tently found in the majority of single subject BOLD activations (19/20 subjects). Significant
BOLD responses were localized to the sensorimotor cortex which includes the postcentral
gyrus (S1, S2), primary and premotor cortex, posterior insula, and deep cerebellum. For the 25
cm/s stimulus condition, BOLD responses were found in the insula in 13/20 subjects. The spa-
tial extent of the evoked BOLD response was significantly dependent on saltatory tactile veloc-
ity with the largest response apparent at 25 cm/s. The probabilistic cytoarchitectonic maps in
SPM Anatomy toolbox v2.2b were used to identify the brain region corresponding to the peak
MNI coordinates from the main effects results [48–51].
Neural encoding of pneumotactile velocity
PLOS ONE | https://doi.org/10.1371/journal.pone.0183532 August 25, 2017 6 / 17

Main effect of various velocity stimuli
The t-contrast results from main effects for 5 cm/s, 25 cm/s, and 65 cm/s were inserted in a
one-way ANOVA within-subjects analysis to evaluate the group main effect of various velocity
stimuli with significance level set to P
unc
<.0001. The result of the group main effect was used
to identify responsive S1, S2 and the somatosensory association areas. Fig 3 shows the BOLD
response of the main effect of velocity in both cortical activation and its coronal view. The
MNI coordinates and F-values of the main effect of the velocity are listed in Table 1. The result
from the one-way ANOVA within-subjects showed BOLD responses in not only contralateral
and ipsilateral cerebral sensorimotor area (S1, S2, primary motor cortex (M1), supplementary
motor cortex (SMA), posterior insula and postcentral gyrus), but also ipsilateral cerebellum.
The peak level of contralateral BOLD response was found in BA3b [MNI (mm) = -47, -20, 58;
F = 56.18], followed by postcentral gyrus [MNI (mm) = -62, -17, 35; F = 28.21]. The highest
level of ipsilateral BOLD response was found in the precentral gyrus [MNI (mm) = 51, 1, 50;
F = 28.99] followed by cerebellum near the dentate nucleus [MNI (mm) = 26, -55, -23;
F = 26.97].
BOLD signal changes and time series
The peak MNI coordinates of left BA3b (-47, -20, 58), BA1 (-62, -17, 35), BA44 (-60, 3, 35),
BA3a (-27, -35, 48), and right Cerebellum (26, -55, -23) were selected from the result of the
main effect for stimulus velocity. The resulting MNI coordinates were estimated with 34%,
39%, 12%, 27% and 98% probability for the left BA3b, left BA1, left BA44, left BA3a, and right
cerebellum, respectively, by using the ANATOMY toolbox v2.2b. Fig 4 shows the BOLD signal
changes and BOLD response time courses among the 20 subjects pooled for each of the 3 con-
ditions compared to rest (no stimulus) in left BA3b, BA1, BA44, BA3a, and right cerebellum
Fig 3. The main effect of velocity stimuli. The main effect of velocity from 20 neurotypical subjects combining 3 different velocities
stimulus (5 cm/s, 25 cm/s, and 65 cm/s). Color-coded evoked BOLD responses at each row indicate brain regions (sagittal, coronal and axial
view) with high F-values. Most of the BOLD responses in Table 1 are represented in this figure.
https://doi.org/10.1371/journal.pone.0183532.g003
Neural encoding of pneumotactile velocity
PLOS ONE | https://doi.org/10.1371/journal.pone.0183532 August 25, 2017 7 / 17

(estimated as the mean of percentage BOLD signal changes across the 20 seconds stimulus
block, P
unc
<.0001). The %BOLD signal changes for each area were calculated by using the
ANATOMY toolbox v2.2b. The largest %BOLD signal changes as a function of saltatory pneu-
motactile velocity were found at left BA3b, followed by left BA44 and BA1. The smallest %
BOLD signal changes in left BA3b were found for the 5 cm/s and no stimulation contrast. The
%BOLD signal change in the left BA3b increased as a function of saltation velocity. The pattern
of BOLD modulation in the left BA1, however, was reversed with greatest BOLD signal change
associated with the 5 cm/s contrast and progressively smaller BOLD signal change at 25 cm/s
and 65 cm/s. The right deep cerebellum, left BA44, and left BA3a showed significant BOLD
signals at the 5 and 25 cm/sec saltation rates, with attenuation of the evoked BOLD response at
the highest velocity of 65 cm/s. The peak BOLD response in these five ROI time series were
found 5 seconds after stimulus onset with the ‘65 cm/s >No stimulus’ contrast showing the
greatest BOLD response in left BA3b. Most of the %BOLD signal change results are generally
consistent with the BOLD time series functions (Fig 4).
One sample t-test (velocities >no stimulus)
The results from one sample t-test in the second-level analysis showed a group result of one-
sided individual velocities compared to the two control conditions (All TAC-Cell pneumatics
OFF and ON). When the individual velocities were compared to the All TAC-Cells OFF
condition (No stimulus) in Fig 5 (the contrasts: 5 cm/s >All TAC-Cells OFF, 25 cm/s >All
TAC-Cells OFF, and 65 cm/s >All TAC-Cells OFF), the contralateral BOLD activations in
sensorimotor cortex were found consistently across most subjects, with the largest spatial
extent and t-values of the evoked BOLD responses at ‘25 cm/s >All TAC-Cells OFF’. MNI
coordinates, t-value, and brain regions are listed in Table 2. Contralateral BOLD responses
localized predominantly to sensorimotor cortex (BA1, BA2, and pre- and postcentral gyrus)
were found in both ‘5 cm/s >No stimulus’ and ‘65 cm/s >No stimulus’ contrasts, whereas
Table 1. Main effect of the velocity MNI coordinates.
MNI Coordinates Cluster-level Peak-level Region
x y z P
FWE-corr
Extent (k
E
) F-value Z P
uncorr
-47 -20 58 0.000 163 56.18 6.82 0.000 L BA3b
51 1 50 0.005 56 28.99 5.47 0.000 R Precentral Gyrus
-62 -17 35 0.000 193 28.21 5.41 0.000 L BA1
26 -55 -23 0.016 39 26.97 5.32 0.000 R Cerebellum
-60 3 35 0.000 293 26.78 5.31 0.000 L BA44
-45 -5 53 26.11 5.26 0.000 L Precentral Gyrus
-55 -2 43 15.61 4.73 0.000 L Precentral Gyrus
-5 1 65 0.000 130 25.01 5.17 0.000 L BA6
-27 -35 48 0.000 106 23.29 5.03 0.000 L BA3a
-35 -35 43 15.61 4.24 0.000 L BA3a
-30 -40 58 15.46 4.22 0.000 L BA2
-50 -37 23 0.056 24 15.40 4.21 0.000 L Superior Temporal Gyrus
56 -35 20 0.097 18 14.99 4.16 0.000 R Superior Temporal Gyrus
One-way ANOVA within-subjects revealed a significant (p (peak-level) <.0001, uncorrected) main effect of the saltatory pneumotactile velocity stimulation.
Cluster-level: The number of activated voxels comprising a cluster. Peak-level: The height of maximum voxel within the cluster, P
FWE-corr
: family-wise error
correction, Extent threshold k >10 voxels, P
uncorr
: uncorrected, BA: Brodmann area, L: Left, R: Right.
https://doi.org/10.1371/journal.pone.0183532.t001
Neural encoding of pneumotactile velocity
PLOS ONE | https://doi.org/10.1371/journal.pone.0183532 August 25, 2017 8 / 17

‘25 cm/s >No stimulus’ contrast evoked significant BOLD responses in BA1, BA43 (a portion
of S2 proximal to the posterior end of the lateral fissure of Sylvius) and postcentral gyrus. The
ipsilateral BOLD responses were found in the inferior parietal lobule (IPL) only at ‘25 cm/s >No
stimulus’ [MNI (mm) = 53, -27, 23; t = 8.22].
Fig 4. BOLD signal changes and BOLD response time courses. Top: The bar graphs show the BOLD signal changes of 3
velocities compared to rest (no stimulus) in Left BA3b, Left BA1, Right Cerebellum, Left BA44 and Left BA3a with SEM
(estimated as the mean of percentage BOLD signal changes across the 20 seconds stimulus block, P
unc
<.0001. Blue and Red
indicate contralateral and ipsilateral to the stimulus, respectively). Bottom: BOLD response time courses corresponding with
each area from the BOLD signal changes (estimated as the average BOLD responses across 20 subjects during the 40
seconds block including stimulus ON and OFF, Blue = 5 cm/s >NO stimulus, Red = 25 cm/s >No stimulus, Green = 65 cm/
s>No stimulus). Y-scales are same for five figures in each row (% BOLD change and BOLD response time courses).
https://doi.org/10.1371/journal.pone.0183532.g004
Fig 5. Group results. Velocities >no stimulus. One sample t-test result on the rendered brain cortical
surface using bspmview (http://www.bobspunt.com/bspmview/) [from the top: (1) 5 cm/s, (2) 25 cm/s, and (3)
65 cm/s >No stimulus, P
unc
<.0001].
https://doi.org/10.1371/journal.pone.0183532.g005
Neural encoding of pneumotactile velocity
PLOS ONE | https://doi.org/10.1371/journal.pone.0183532 August 25, 2017 9 / 17

One sample t-test (velocities >all on)
The individual velocities were compared to the All TAC-Cells ON condition as shown in Fig 6.
The contralateral BOLD responses in sensorimotor cortex (BA3, BA6 and pre- and postcentral
gyrus) were found for the three contrasts: ‘5 cm/s >All TAC-Cells ON’, ‘25 cm/s >All TAC-
Table 2. One sample t-test results. Velocities >no stimulus.
MNI Coordinates t-value P
uncorr
Region
x y z
5cm/s >No stimulus -65 -20 43 8.70 0.000 L BA1
-52 -17 20 8.64 0.000 L Postcentral Gyrus
-57 -22 50 6.80 0.000 L BA1
-37 -35 45 5.89 0.000 L BA2
-50 -5 8 5.66 0.000 L Precentral Gyrus
-45 -5 15 5.04 0.000 L Rolandic Operculum
25cm/s >No stimulus -50 -25 55 11.92 0.000 L BA1
-47 -17 18 10.83 0.000 L BA43
-55 -15 20 10.31 0.000 L Postcentral Gyrus
53 -27 23 8.22 0.000 R Inferior Parietal Lobule
65cm/s >No stimulus -52 -22 55 9.40 0.000 L BA1
-47 -17 20 6.98 0.000 L Rolandic Operculum (OP3)
One sample t-test revealed a significant (p<0.0001, uncorrected) BOLD response of one-sided individual velocities compared to all TAC-Cells OFF (No
stimulus). BA: Brodmann area, L: Left, R: Right, P
uncorr
: uncorrected.
https://doi.org/10.1371/journal.pone.0183532.t002
Fig 6. Group results. Velocities >all-ON. One sample t-test result on the rendered brain cortical surface
using bspmview (http://www.bobspunt.com/bspmview/) [from the top: (1) 5 cm/s, (2) 25 cm/s, and (3)
65 cm/s >All-ON, P
unc
<.0001].
https://doi.org/10.1371/journal.pone.0183532.g006
Neural encoding of pneumotactile velocity
PLOS ONE | https://doi.org/10.1371/journal.pone.0183532 August 25, 2017 10 / 17

Cells ON’, and ‘65 cm/s >All TAC-Cells ON’, whereas the ipsilateral BOLD activations in
BA2 were seen only at ‘5 cm/s >All TAC-Cells ON’ [MNI (mm) = 33, -37, 45; t = 6.35]. As
shown in Table 3, the peak t-value was observed at ‘5 cm/s >All TAC-Cells ON’ (t = 9.18)
while relatively small BOLD responses were found at the highest velocity condition
(65 cm/s >All TAC-Cells ON). The spatial extent of BOLD responses at ‘5 cm/s >All TAC-
Cells ON’ and ‘25 cm/s >All TAC-Cells ON’ were larger than the highest velocity contrast.
Discussion
In this study, we used a new saltatory pneumotactile stimulus modality programmed at 3 dif-
ferent velocities on the glabrous hand to map the evoked hemodynamic BOLD response in
cortical and subcortical somatosensory areas using fMRI methods. Overall, the BOLD main
effect for saltatory pneumotactile velocity was localized to several loci involving contralateral
and bilateral cerebral cortex, and ipsilateral cerebellum. This elaborate network extends previ-
ous observations based on fMRI and MEG of pneumotactile encoding of single channel pulse
train inputs (non-saltatory) which found principal dipoles localized to S1, S2 and PPC in both
contralateral and bilateral cerebral sensorimotor cortex [52,53]. Additionally, the cerebral
responses in S1 and PPC are generally consistent with the findings from our previous MEG
studies using the first and second generation of TAC-Cells (19.3 mm ID, and 6 mm ID, respec-
tively) developed in our laboratory [43,44,53]. We also found significant ipsilateral BOLD
responses in deep cerebellum which was reported in previous fMRI and PET studies using the
brush and the foam-tipped motor to create the movement of the tactile stimulus (tickling) on
the palm [17,19,54]. Moreover, our pneumotactile saltatory stimulation on the glabrous hand
produced the largest spatial extent of the evoked BOLD responses at ‘25 cm/s >No stimulus’,
which corresponds closely to the zenith of perceptual capacity for tactile traverse velocity (5 to
30 cm/sec) revealed by human skin psychophysics using a traversing soft brush. Although not
a continuous input, the highly effective nature of saltatory pneumotactile inputs on the gla-
brous hand at 5 cm/s and 25 cm/s shares features of the optimal operating range for discrimi-
nation of velocity observed psychophysical studies using continuous brush stimulation applied
to the glabrous skin in humans [55]. Parallels in optimal stimulus of perceptual velocity can be
Table 3. One sample t-test results. Velocities >all-ON.
MNI Coordinates t-value P
uncorr
Region
x y z
5cm/s >All-ON -67 -17 33 9.18 0.000 L Superior Temporal Gyrus
33 -37 45 6.35 0.000 R BA2
-40 -35 48 5.68 0.000 L BA2
-32 -32 45 5.57 0.000 L BA3a
-30 -10 58 4.99 0.000 L BA6
-25 -7 50 4.71 0.000 L BA6
25cm/s >All-ON -60 -17 40 6.96 0.000 L BA1
-57 -2 45 6.66 0.000 L BA6
-62 -15 30 6.23 0.000 L BA1
-37 -10 65 5.26 0.000 L Precentral Gyrus
65cm/s >All-ON -47 -20 58 5.75 0.000 L BA3b
One sample t-test revealed a significant (p<0.0001, uncorrected) BOLD response of one-sided individual velocities compared to all TAC-Cells ON. BA:
Brodmann area, L: Left, R: Right, P
uncorr
: uncorrected.
https://doi.org/10.1371/journal.pone.0183532.t003
Neural encoding of pneumotactile velocity
PLOS ONE | https://doi.org/10.1371/journal.pone.0183532 August 25, 2017 11 / 17

drawn from the results of single-unit recordings in non-human primate somatosensory cortex
during continuous skin brushing [4,38,56].
We have demonstrated regional neural activation using pneumotactile saltatory stimulation
via TAC-Cells in both contralateral (20/20 subjects) and ipsilateral somatosensory cortex (6/20
subjects) in in an effort to better understand how various stimulus velocity profiles influence
the spatial extent and ROI modulation of brain networks in neurotypical adults. Our results
showed that the contralateral BOLD responses were found at sensorimotor cortex (S1, S2, M1,
SMA, pre- and postcentral gyrus) across the most subjects (19/20 subjects), whereas the ipsilat-
eral BOLD activations were limited to S1, S2, and deep cerebellum only in 6/20 subjects. The
predominantly contralateral BOLD response in the hand representation of the sensorimotor
cortex is consistent with human fMRI studies using electrical and laser stimulation [57,58]. In
addition, our finding of a significant ipsilateral BOLD response in the ipsilateral cerebellum is
consistent with a previous human PET study using finger movements to create tactile stimula-
tion [59]. Additionally, significant BOLD response in deep cerebellum at 25 cm/sec is consis-
tent with its presumed role in sensory information processing for monitoring and optimizing
movement using sensory proprioceptive feedback information [60].
Our results also show that the spatial extent of BOLD responses increased dramatically
when stimulus velocity was increased from 5 cm/s to 25 cm/s, and significantly decreased or
‘funneled’ at the highest velocity of 65 cm/s. The %BOLD response change in contralateral S1
(BA3b), showed a robust increase as a function of increasing velocity which is consistent with
previous fMRI studies using tactile and non-painful electrical stimulation of the median nerve
[42,61,62]. The evoked response characteristics associated with moving tactile stimulation
suggest that the fast-adapting (FA) mechanoreceptors, which are heavily concentrated in the
glabrous hand, are sensitive to tactile stimulation velocity [63]. Further, we discovered the ipsi-
lateral BOLD signal in the IPL at ‘25 cm/s >No stimulus’ contrast. The IPL has been hypothe-
sized to play a role in sensorimotor integration [64,65].
The TAC-Cells developed in our laboratory are safe, non-invasive, simple, rapidly applica-
ble to the skin, fully compatible with MRI/MEG, produce no stimulus artifact, achieve normal
‘physiologic’ recruitment order of primary mechanosensitive afferents, avoid the potential
risks associated with direct-current stimulation methods, and are well tolerated by subjects
across the lifespan from infancy through adulthood. Most previous studies were limited to
study the median nerve using electrical current stimulation, which noted increasing somato-
sensory thresholds during the stimulation in healthy human subjects [66]. The TAC-Cells rep-
resent a natural mode of tactile stimulation via a small pneumatically charged capsule, which
can be placed on virtually any skin surface of the body, including the glabrous hand and face.
Our multi-channel pneumotactile stimulus array control system (GALILEO) can be pro-
grammed to control pulse duration and rise/fall times, relative timing between individual
channels or cells to create unique velocity and direction trajectories over the skin, stimulus
block or event-related design (continuous, random, random-balanced), and various triggering
modes which are essential for task- or stimulus-related fMRI experiments.
Results of this study have generated new information on the spatiotemporal features of sal-
tatory tactile velocity encoding originating from Aβmechanoreceptors in the glabrous hand
projecting along the medial lemniscus to cerebral and deep cerebellar somatosensory represen-
tations in neurotypical adults. Moreover, this work is expected to inform future investigations
whose goal is to develop new approaches to motor rehabilitation through somatosensory neu-
rotherapeutics to improve sensorimotor function in individuals who have sustained cerebro-
vascular stroke or traumatic brain injury. Although the current generation of 3T fMRI
scanners provides relatively high spatial resolution (*2 mm), the temporal resolution is lim-
ited (seconds) due to intrinsic properties of the hemodynamic response [67]. The use of a
Neural encoding of pneumotactile velocity
PLOS ONE | https://doi.org/10.1371/journal.pone.0183532 August 25, 2017 12 / 17

multiband echoplanar sequence can reduce the TR from 2.5 seconds to 1.0 second, thus offer-
ing some improvement in BOLD time series modeling. In addition, 7T fMRI scanners could
be applied which can produce very high resolution functional data (less than 1mm spatial reso-
lution) [68,69]. Multimodal combination of fMRI and EEG, or co-registration studies using
SQUID-based superconducting MEG, or the rapidly evolving technology known as atomic
(AM) or optically-pumped magnetometers (OPM) would potentially yield the best available
spatial and temporal resolution to reveal the dynamics of the human somatosensory brain.
Moreover, a detailed analysis of the BOLD response time series in other sensorimotor areas
(e.g. BA1, BA2, BA4, and cerebellum) could be employed to develop a model of functional
brain connectivity as a function of stimulus traverse velocity.
In summary, we found that the TAC-Cell pneumotactile stimulus array delivered at 3 dif-
ferent velocities on the glabrous hand was highly effective at evoking and modulating BOLD
responses among 5 ROIs in primary and secondary sensorimotor cortices and deep cerebel-
lum. The strong dependence of %BOLD change found in the present study maps well to
known psychophysical and electrophysiological findings in animal and human models and
shows potential relevance for motor control of the hand. The spatial extent of BOLD responses
was also significantly dependent on the velocity of tactile stimuli.
Acknowledgments
This work was supported by the internal grant from Barkley Trust Foundation (SM Barlow—
PI). There was no additional external funding received for this study.
Author Contributions
Conceptualization: Hyuntaek Oh, Rebecca Custead, Steven Barlow.
Data curation: Hyuntaek Oh, Rebecca Custead, Yingying Wang, Steven Barlow.
Formal analysis: Hyuntaek Oh, Rebecca Custead, Yingying Wang, Steven Barlow.
Funding acquisition: Steven Barlow.
Investigation: Hyuntaek Oh, Steven Barlow.
Methodology: Hyuntaek Oh, Yingying Wang, Steven Barlow.
Project administration: Hyuntaek Oh, Steven Barlow.
Resources: Hyuntaek Oh, Steven Barlow.
Software: Hyuntaek Oh, Steven Barlow.
Supervision: Hyuntaek Oh, Steven Barlow.
Validation: Hyuntaek Oh, Yingying Wang, Steven Barlow.
Visualization: Hyuntaek Oh.
Writing – original draft: Hyuntaek Oh, Steven Barlow.
Writing – review & editing: Hyuntaek Oh, Steven Barlow.
References
1. Buonomano DV, Merzenich MM. Cortical plasticity: from synapses to maps. Annual Review of Neuro-
science. 1998; 21(1):149–186. https://doi.org/10.1146/annurev.neuro.21.1.149 PMID: 9530495
Neural encoding of pneumotactile velocity
PLOS ONE | https://doi.org/10.1371/journal.pone.0183532 August 25, 2017 13 / 17

2. Khaslavskaia S, Ladouceur M, Sinkjaer T. Increase in tibialis anterior motor cortex excitability following
repetitive electrical stimulation of the common peroneal nerve. Experimental Brain Research. 2002;
145(3):309–315. https://doi.org/10.1007/s00221-002-1094-9 PMID: 12136380
3. Dreyer D, Hollins M, Whitsel B. Factors influencing cutaneous directional sensitivity. Sensory Pro-
cesses. 1978; 2(2):71–79. PMID: 715470
4. Whitsel B, Franzen O, Dreyer D, Hollins M, Young M, Essick G, et al. Dependence of subjective tra-
verse length on velocity of moving tactile stimuli. Somatosensory Research. 1986; 3(3):185–196.
https://doi.org/10.3109/07367228609144583 PMID: 3749661
5. Olausson H, Norrsell U. Observations on human tactile directional sensibility. The Journal of Physiol-
ogy. 1993; 464(1):545–559. https://doi.org/10.1113/jphysiol.1993.sp019650 PMID: 8229817
6. Custead R, Oh H, Rosner AO, Barlow S. Adaptation of the cortical somatosensory evoked potential fol-
lowing pulsed pneumatic stimulation of the lower face in adults. Brain Research. 2015; 1622:81–90.
https://doi.org/10.1016/j.brainres.2015.06.025 PMID: 26119917
7. Wienbruch C, Candia V, Svensson J, Kleiser R, Kollias SS. A portable and low-cost fMRI compatible
pneumatic system for the investigation of the somatosensensory system in clinical and research envi-
ronments. Neuroscience Letters. 2006; 398(3):183–188. https://doi.org/10.1016/j.neulet.2006.01.025
PMID: 16469438
8. Huang RS, Sereno MI. Dodecapus: an MR-compatible system for somatosensory stimulation. Neuro-
Image. 2007; 34(3):1060–1073. https://doi.org/10.1016/j.neuroimage.2006.10.024 PMID: 17182259
9. Bertolasi L, Priori A, Tinazzi M, Bertasi V, Rothwell JC. Inhibitory action of forearm flexor muscle affer-
ents on corticospinal outputs to antagonist muscles in humans. The Journal of Physiology. 1998;
511(3):947–956. https://doi.org/10.1111/j.1469-7793.1998.947bg.x PMID: 9714872
10. Ridding M, Brouwer B, Miles T, Pitcher J, Thompson P. Changes in muscle responses to stimulation of
the motor cortex induced by peripheral nerve stimulation in human subjects. Experimental Brain
Research. 2000; 131(1):135–143. https://doi.org/10.1007/s002219900269 PMID: 10759179
11. Nitsche MA, Paulus W. Excitability changes induced in the human motor cortex by weak transcranial
direct current stimulation. The Journal of Physiology. 2000; 527(3):633–639. https://doi.org/10.1111/j.
1469-7793.2000.t01-1-00633.x PMID: 10990547
12. Tommerdahl M, Simons SB, Chiu JS, Favorov O, Whitsel BL. Ipsilateral input modifies the primary
somatosensory cortex response to contralateral skin flutter. Journal of Neuroscience. 2006; 26(22):
5970–5977. https://doi.org/10.1523/JNEUROSCI.5270-05.2006 PMID: 16738239
13. Chen TL, Babiloni C, Ferretti A, Perrucci MG, Romani GL, Rossini PM,et al. Human secondary somato-
sensory cortex is involved in the processing of somatosensory rare stimuli: an fMRI study. NeuroImage.
2008; 40(4):1765–1771. https://doi.org/10.1016/j.neuroimage.2008.01.020 PMID: 18329293
14. Rozzi S, Calzavara R, Belmalih A, Borra E, Gregoriou GG, Matelli M, et al. Cortical connections of the
inferior parietal cortical convexity of the macaque monkey. Cerebral Cortex. 2005; 16(10):1389–1417.
https://doi.org/10.1093/cercor/bhj076 PMID: 16306322
15. Karkhanis AN, Heider B, Silva FM, Siegel RM. Spatial effects of shifting prisms on properties of poste-
rior parietal cortex neurons. The Journal of Physiology. 2014; 592(16):3625–3646. https://doi.org/10.
1113/jphysiol.2014.270942 PMID: 24928956
16. Xing J, Andersen RA. Models of the posterior parietal cortex which perform multimodal integration and
represent space in several coordinate frames. Journal of Cognitive Neuroscience. 2000; 12(4):
601–614. https://doi.org/10.1162/089892900562363 PMID: 10936913
17. Blakemore SJ, Wolpert DM, Frith CD. The cerebellum contributes to somatosensory cortical activity
during self-produced tactile stimulation. NeuroImage. 1999; 10(4):448–459. https://doi.org/10.1006/
nimg.1999.0478 PMID: 10493902
18. Blakemore SJ, Sirigu A. Action prediction in the cerebellum and in the parietal lobe. Experimental Brain
Research. 2003; 153(2):239–245. https://doi.org/10.1007/s00221-003-1597-z PMID: 12955381
19. Blakemore SJ, Frith CD, Wolpert DM. The cerebellum is involved in predicting the sensory conse-
quences of action. NeuroReport. 2001; 12(9):1879–1884. https://doi.org/10.1097/00001756-
200107030-00023 PMID: 11435916
20. Kawato M, Kuroda T, Imamizu H, Nakano E, Miyauchi S, Yoshioka T. Internal forward models in the cer-
ebellum: fMRI study on grip force and load force coupling. Progress in Brain Research. 2003; 142:
171–188. https://doi.org/10.1016/S0079-6123(03)42013-X PMID: 12693261
21. Bursztyn LL, Ganesh G, Imamizu H, Kawato M, Flanagan JR. Neural correlates of internal-model load-
ing. Current Biology. 2006; 16(24):2440–2445. https://doi.org/10.1016/j.cub.2006.10.051 PMID:
17174919
Neural encoding of pneumotactile velocity
PLOS ONE | https://doi.org/10.1371/journal.pone.0183532 August 25, 2017 14 / 17

22. Johansson RS, Flanagan JR. Coding and use of tactile signals from the fingertips in object manipulation
tasks. Nature Reviews Neuroscience. 2009; 10(5):345. https://doi.org/10.1038/nrn2621 PMID:
19352402
23. Small S, Hlustik P, Noll D, Genovese C, Solodkin A. Cerebellar hemispheric activation ipsilateral to the
paretic hand correlates with functional recovery after stroke. Brain. 2002; 125(7):1544–1557. https://
doi.org/10.1093/brain/awf148 PMID: 12077004
24. Tseng Yw, Diedrichsen J, Krakauer JW, Shadmehr R, Bastian AJ. Sensory prediction errors drive cere-
bellum-dependent adaptation of reaching. Journal of Neurophysiology. 2007; 98(1):54–62. https://doi.
org/10.1152/jn.00266.2007 PMID: 17507504
25. Lin YY, Kajola M. Neuromagnetic somatosensory responses to natural moving tactile stimulation. Cana-
dian Journal of Neurological Sciences. 2003; 30(1):31–35. https://doi.org/10.1017/
S0317167100002407 PMID: 12619781
26. Essick G, Franzen O, Whitsel B. Discrimination and scaling of velocity of stimulus motion across the
skin. Somatosensory & Motor Research. 1988; 6(1):21–40. https://doi.org/10.3109/
08990228809144639
27. Lo
¨ken LS, Evert M, Wessberg J. Pleasantness of touch in human glabrous and hairy skin: order effects
on affective ratings. Brain Research. 2011; 1417:9–15. https://doi.org/10.1016/j.brainres.2011.08.011
PMID: 21907328
28. Bjo
¨rnsdotter M, Olausson H. Vicarious responses to social touch in posterior insular cortex are tuned to
pleasant caressing speeds. Journal of Neuroscience. 2011; 31(26):9554–9562. https://doi.org/10.1523/
JNEUROSCI.0397-11.2011 PMID: 21715620
29. Ackerley R, Hassan E, Curran A, Wessberg J, Olausson H, McGlone F. An fMRI study on cortical
responses during active self-touch and passive touch from others. Frontiers in Behavioral Neurosci-
ence. 2012; 6. https://doi.org/10.3389/fnbeh.2012.00051 PMID: 22891054
30. McGlone F, Reilly D. The cutaneous sensory system. Neuroscience & Biobehavioral Reviews. 2010;
34(2):148–159. https://doi.org/10.1016/j.neubiorev.2009.08.004
31. McGlone F, Wessberg J, Olausson H. Discriminative and affective touch: sensing and feeling. Neuron.
2014; 82(4):737–755. https://doi.org/10.1016/j.neuron.2014.05.001 PMID: 24853935
32. Pei YC, Bensmaia SJ. The neural basis of tactile motion perception. Journal of Neurophysiology. 2014;
112(12):3023–3032. https://doi.org/10.1152/jn.00391.2014 PMID: 25253479
33. Dallmann CJ, Ernst MO, Moscatelli A. The role of vibration in tactile speed perception. Journal of Neuro-
physiology. 2015; 114(6):3131–3139. https://doi.org/10.1152/jn.00621.2015 PMID: 26424580
34. Iggo A. Somatosensory system. vol. 2. Springer Science & Business Media; 2012.
35. Weibull A, Bjo
¨rkman A, Hall H, Rose
´n B, Lundborg G, Svensson J. Optimizing the mapping of finger
areas in primary somatosensory cortex using functional MRI. Magnetic Resonance Imaging. 2008;
26(10):1342–1351. https://doi.org/10.1016/j.mri.2008.04.007 PMID: 18550314
36. Martuzzi R, Zwaag W, Farthouat J, Gruetter R, Blanke O. Human finger somatotopy in areas 3b, 1, and
2: a 7T fMRI study using a natural stimulus. Human Brain Mapping. 2014; 35(1):213–226. https://doi.
org/10.1002/hbm.22172 PMID: 22965769
37. Schweisfurth MA, Frahm J, Schweizer R. Individual fMRI maps of all phalanges and digit bases of all fin-
gers in human primary somatosensory cortex. Frontiers in Human Neuroscience. 2014; 8. https://doi.
org/10.3389/fnhum.2014.00658
38. Whitsel B, Dreyer D, Hollins M. Representation of moving stimuli by somatosensory neurons. Federa-
tion Proceedings. 1978; 37(9):2223–2227. PMID: 95974
39. Keyson DV, Houtsma AJ. Directional sensitivity to a tactile point stimulus moving across the fingerpad.
Attention, Perception, & Psychophysics. 1995; 57(5):738–744. https://doi.org/10.3758/BF03213278
40. Olausson H, Lamarre Y, Backlund H, Morin C, Wallin B, Starck G, et al. Unmyelinated tactile afferents
signal touch and project to insular cortex. Nature Neuroscience. 2002; 5(9):900. https://doi.org/10.1038/
nn896 PMID: 12145636
41. Gleeson BT, Horschel SK, Provancher WR. Perception of direction for applied tangential skin displace-
ment: Effects of speed, displacement, and repetition. IEEE transactions on haptics. 2010; 3(3):
177–188. https://doi.org/10.1109/TOH.2010.20 PMID: 27788072
42. Hlushchuk Y, Simões-Franklin C, Nangini C, Hari R. Stimulus-Rate Sensitivity Discerns Area 3b of the
Human Primary Somatosensory Cortex. PloS one. 2015; 10(5):e0128462. https://doi.org/10.1371/
journal.pone.0128462 PMID: 26020639
43. Venkatesan L, Barlow S, Popescu M, Popescu A, Auer ET. TAC-Cell inputs to human hand and lip
induce short-term adaptation of the primary somatosensory cortex. Brain Research. 2010; 1348:63–70.
https://doi.org/10.1016/j.brainres.2010.06.015 PMID: 20550944
Neural encoding of pneumotactile velocity
PLOS ONE | https://doi.org/10.1371/journal.pone.0183532 August 25, 2017 15 / 17

44. Venkatesan L, Barlow SM, Popescu M, Popescu A. Integrated approach for studying adaptation mech-
anisms in the human somatosensory cortical network. Experimental Brain Research. 2014; 232(11):
3545–3554. https://doi.org/10.1007/s00221-014-4043-5 PMID: 25059913
45. Verrillo RT, Bolanowski SJ Jr. The effects of skin temperature on the psychophysical responses to
vibration on glabrous and hairy skin. The Journal of the Acoustical Society of America. 1986; 80(2):
528–532. https://doi.org/10.1121/1.394047 PMID: 3745684
46. Friston KJ, Holmes AP, Poline J, Grasby P, Williams S, Frackowiak RS, et al. Analysis of fMRI time-
series revisited. NeuroImage. 1995; 2(1):45–53. https://doi.org/10.1006/nimg.1995.1007 PMID:
9343589
47. Friston KJ, Stephan KE, Lund TE, Morcom A, Kiebel S. Mixed-effects and fMRI studies. NeuroImage.
2005; 24(1):244–252. https://doi.org/10.1016/j.neuroimage.2004.08.055 PMID: 15588616
48. Eickhoff SB, Stephan KE, Mohlberg H, Grefkes C, Fink GR, Amunts K, et al. A new SPM toolbox for
combining probabilistic cytoarchitectonic maps and functional imaging data. NeuroImage. 2005;
25(4):1325–1335. https://doi.org/10.1016/j.neuroimage.2004.12.034 PMID: 15850749
49. Eickhoff SB, Heim S, Zilles K, Amunts K. Testing anatomically specified hypotheses in functional imag-
ing using cytoarchitectonic maps. NeuroImage. 2006; 32(2):570–582. https://doi.org/10.1016/j.
neuroimage.2006.04.204 PMID: 16781166
50. Eickhoff SB, Paus T, Caspers S, Grosbras MH, Evans AC, Zilles K, et al. Assignment of functional acti-
vations to probabilistic cytoarchitectonic areas revisited. NeuroImage. 2007; 36(3):511–521. https://doi.
org/10.1016/j.neuroimage.2007.03.060 PMID: 17499520
51. Papadelis C, Eickhoff SB, Zilles K, Ioannides AA. BA3b and BA1 activate in a serial fashion after
median nerve stimulation: direct evidence from combining source analysis of evoked fields and
cytoarchitectonic probabilistic maps. NeuroImage. 2011; 54(1):60–73. https://doi.org/10.1016/j.
neuroimage.2010.07.054 PMID: 20691793
52. Brodoehl S, Klingner C, Stieglitz K, Witte OW. Age-related changes in the somatosensory processing of
tactile stimulation?An fMRI study. Behavioural Brain Research. 2013; 238:259–264. https://doi.org/10.
1016/j.bbr.2012.10.038 PMID: 23123141
53. Popescu EA, Barlow SM, Venkatesan L, Wang J, Popescu M. Adaptive changes in the neuromagnetic
response of the primary and association somatosensory areas following repetitive tactile hand stimula-
tion in humans. Human Brain Mapping. 2013; 34(6):1415–1426. https://doi.org/10.1002/hbm.21519
PMID: 22331631
54. Hagen MC, Franze
´n O, McGlone F, Essick G, Dancer C, Pardo JV. Tactile motion activates the human
middle temporal/V5 (MT/V5) complex. European Journal of Neuroscience. 2002; 16(5):957–964.
https://doi.org/10.1046/j.1460-9568.2002.02139.x PMID: 12372032
55. McGlone F, Olausson H, Boyle J, Jones-Gotman M, Dancer C, Guest S, et al. Touching and feeling: dif-
ferences in pleasant touch processing between glabrous and hairy skin in humans. European Journal of
Neuroscience. 2012; 35(11):1782–1788. https://doi.org/10.1111/j.1460-9568.2012.08092.x PMID:
22594914
56. Lee CJ, Whitsel BL. Mechanisms underlying somatosensory cortical dynamics: I. In vivo studies. Cere-
bral Cortex. 1992; 2(2):81–106. https://doi.org/10.1093/cercor/2.2.107 PMID: 1633418
57. Backes W, Mess W, van Kranen-Mastenbroek V, Reulen J. Somatosensory cortex responses to
median nerve stimulation: fMRI effects of current amplitude and selective attention. Clinical Neurophysi-
ology. 2000; 111(10):1738–1744. https://doi.org/10.1016/S1388-2457(00)00420-X PMID: 11018487
58. Bornho
¨vd K, Quante M, Glauche V, Bromm B, Weiller C, Bu¨chel C. Painful stimuli evoke different stimu-
lus–response functions in the amygdala, prefrontal, insula and somatosensory cortex: a single-trial
fMRI study. Brain. 2002; 125(6):1326–1336. https://doi.org/10.1093/brain/awf137 PMID: 12023321
59. Fox PT, Raichle ME, Thach WT. Functional mapping of the human cerebellum with positron emission
tomography. Proceedings of the National Academy of Sciences. 1985; 82(21):7462–7466. https://doi.
org/10.1073/pnas.82.21.7462
60. Gazzola V, Keysers C. The observation and execution of actions share motor and somatosensory vox-
els in all tested subjects: single-subject analyses of unsmoothed fMRI data. Cerebral Cortex. 2008;
19(6):1239–1255. https://doi.org/10.1093/cercor/bhn181 PMID: 19020203
61. Kampe KK, Jones RA, Auer DP. Frequency dependence of the functional MRI response after electrical
median nerve stimulation. Human Brain Mapping. 2000; 9(2):106–114. https://doi.org/10.1002/(SICI)
1097-0193(200002)9:2%3C106::AID-HBM5%3E3.0.CO;2-Y PMID: 10680767
62. Ferretti A, Babiloni C, Arienzo D, Del Gratta C, Rossini PM, Tartaro A, et al. Cortical brain responses
during passive nonpainful median nerve stimulation at low frequencies (0.5–4 Hz): an fMRI study.
Human Brain Mapping. 2007; 28(7):645–653. https://doi.org/10.1002/hbm.20292 PMID: 17094120
Neural encoding of pneumotactile velocity
PLOS ONE | https://doi.org/10.1371/journal.pone.0183532 August 25, 2017 16 / 17

63. Essick GK, Edin BB. Receptor encoding of moving tactile stimuli in humans. II. The mean response of
individual low-threshold mechanoreceptors to motion across the receptive field. Journal of Neurosci-
ence. 1995; 15(1):848–864. PMID: 7823185
64. Kitada R, Kochiyama T, Hashimoto T, Naito E, Matsumura M. Moving tactile stimuli of fingers are inte-
grated in the intraparietal and inferior parietal cortices. NeuroReport. 2003; 14(5):719–724. https://doi.
org/10.1097/00001756-200304150-00012 PMID: 12692470
65. Caspers S, Schleicher A, Bacha-Trams M, Palomero-Gallagher N, Amunts K, Zilles K. Organization of
the human inferior parietal lobule based on receptor architectonics. Cerebral Cortex. 2012; 23(3):
615–628. https://doi.org/10.1093/cercor/bhs048 PMID: 22375016
66. Dean J, Bowsher D, Johnson MI. The effects of unilateral transcutaneous electrical nerve stimulation of
the median nerve on bilateral somatosensory thresholds. Clinical Physiology and functional Imaging.
2006; 26(5):314–318. https://doi.org/10.1111/j.1475-097X.2006.00689.x PMID: 16939510
67. Kim SG, Richter W, Ug
ˇurbil K. Limitations of temporal resolution in functional MRI. Magnetic Resonance
in Medicine. 1997; 37(4):631–636. https://doi.org/10.1002/mrm.1910370427 PMID: 9094089
68. Sanchez-Panchuelo RM, Besle J, Beckett A, Bowtell R, Schluppeck D, Francis S. Within-digit functional
parcellation of Brodmann areas of the human primary somatosensory cortex using functional magnetic
resonance imaging at 7 tesla. Journal of Neuroscience. 2012; 32(45):15815–15822. https://doi.org/10.
1523/JNEUROSCI.2501-12.2012 PMID: 23136420
69. Besle J, Sa
´nchez-Panchuelo RM, Bowtell R, Francis S, Schluppeck D. Single-subject fMRI mapping at
7 T of the representation of fingertips in S1: a comparison of event-related and phase-encoding designs.
Journal of Neurophysiology. 2013; 109(9):2293–2305. https://doi.org/10.1152/jn.00499.2012 PMID:
23427300
Neural encoding of pneumotactile velocity
PLOS ONE | https://doi.org/10.1371/journal.pone.0183532 August 25, 2017 17 / 17