Functional architecture of the somatosensory

homunculus detected by electrostimulation

Franck-Emmanuel Roux, Imène Djidjeli, Jean-Baptiste Durand

To cite this version:

Franck-Emmanuel Roux, Imène Djidjeli, Jean-Baptiste Durand. Functional architecture of the so-
matosensory homunculus detected by electrostimulation. The Journal of Physiology, 2018, 596 (5),
pp.941-956. �10.1113/jp275243�. �hal-02341978�

HAL Id: hal-02341978

https://hal.science/hal-02341978v1
Submitted on 28 Apr 2022

HAL is a multi-disciplinary open access L’archive ouverte pluridisciplinaire HAL, est

archive for the deposit and dissemination of sci- destinée au dépôt et à la diffusion de documents
entific research documents, whether they are pub- scientifiques de niveau recherche, publiés ou non,
lished or not. The documents may come from émanant des établissements d’enseignement et de
teaching and research institutions in France or recherche français ou étrangers, des laboratoires
abroad, or from public or private research centers. publics ou privés.
 J Physiol 000.00 (2018) pp 1–16 1

Functional architecture of the somatosensory homunculus

detected by electrostimulation
Franck-Emmanuel Roux1,2 , Imène Djidjeli1,2 and Jean-Baptiste Durand1
1
CNRS (CERCO) UMR Unité 5549, Université Paul Sabatier, Toulouse, 31059, France
2
Pôle NeuroSciences (Neurochirurgie), Centres Hospitalo-Universitaires, Toulouse, 31059, France

Edited by: Janet Taylor & Richard Carson

Key points
r We performed a prospective electrostimulation study, based on 50 operated intact patients, to
acquire accurate MNI coordinates of the functional areas of the somatosensory homunculus.
r In the contralateral BA1, the hand representation displayed not only medial-to-lateral,
The Journal of Physiology

little-finger-to-thumb, but also rostral-to-caudal discrete somatotopy, with the tip of each
finger located more caudally than the proximal phalanx.
r The analysis of the MNI body coordinates showed rare inter-individual variations in the
medial-to-lateral somatotopic organization in these patients with rather different intensity
thresholds needed to elicit sensations in different body parts.
r We found some similarities but also substantial differences with the previous, seminal works
of Penfield and his colleagues.
r We propose a new drawing of the human somatosensory homunculus according to MNI space.

Abstract In this prospective electrostimulation study, based on 50 operated patients with no

sensory deficit and no brain lesion in the postcentral gyrus, we acquired coordinates in the
standard MNI space of the functional areas of the somatosensory homunculus. The 3D brain
volume of each patient was normalized to that space to obtain the MNI coordinates of the
stimulation site locations. For 647 sites stimulated on Brodmann Area 1 (and 1025 in gyri
nearby), 258 positive points for somatosensory response (40%) were found in the postcentral
gyrus. In the contralateral BA1, the hand representation displayed not only medial-to-lateral and
little-finger-to-thumb somatotopy, but also rostral-to-caudal discrete somatotopy, with the tip
of each finger located more caudally than the proximal phalanx. We detected a medial-to-lateral,
tip-to-base tongue organization but no rostral-to-caudal functional organization. The analysis
of the MNI body coordinates showed rare inter-individual variations in the medial-to-lateral
somatotopic organization in these patients with intact somatosensory cortex. Positive stimulations
were detected through the ‘on/off’ outbreak effect and discriminative touch sensations were the
sensations reported almost exclusively by all patients during stimulation. Mean hand (2.39 mA)
and tongue (2.60 mA) positive intensity thresholds were lower (P < 0.05) than the intensities
required to elicit sensations in the other parts of the body. Unlike the previous, seminal works
of Penfield and colleagues, we detected no sensations such as sense of movement or desire to
move, no somatosensory responses outside the postcentral gyrus, and no bilateral responses
for face/tongue stimulations. We propose a rationalization of the standard drawing of the
somatosensory homunculus according to MNI space.

(Received 12 September 2017; accepted after revision 18 December 2017; first published online 29 December 2017)
Corresponding author F.-E. Roux: CNRS (CERCO) UMR Unité 5549 and Service de Neurochirurgie, Hôpital Purpan,
F-31059 Toulouse, France. Email: [email protected]


C 2017 The Authors. The Journal of Physiology 
C 2017 The Physiological Society DOI: 10.1113/JP275243
2 F.-E. Roux and others J Physiol 000.00

Introduction Methods
The primary somatosensory cortex can be divided Ethical approval
into four distinct cytoarchitectonic areas (Powell &
Mountcastle, 1959; Jones et al. 1978), named Brodmann Patients less than 18 years old were excluded. The National
areas (BA) 3a, 3b, 1, and 2 (Brodmann 1909; Vogt & Consultative Committee of INSERM (Institut National de
Vogt, 1919), areas 3a and 3b being located within the la Santé et de la Recherche Médicale) gave its approval
central sulcus, area 1 in the crown of the postcentral for the storage of patients’ data and preservation of
gyrus, and area 2 more caudally in the postcentral sulcus their anonymity. To preserve patient privacy, this study
(White et al. 1997). Animal studies have shown that BA was not registered in a publicly accessible database.
3a receives inputs from muscles and joints, whereas BA However, the study conformed to the standards set by
3b and 1 receive from the skin, and BA 2 can combine the Declaration of Helsinki. All the patients and their
skin and proprioceptive information (Kandel et al. 2013). families gave their informed consent for a study of the
This information progresses hierarchically in a rostro- functional areas by direct brain mapping and each chart
caudal direction (Iwamura et al. 1993) with receptive fields was discussed pre-operatively in a surgical staff meeting
increasing in size from area 3b to area 2 (Gardner, 1988). with different neurosurgeons. Once 50 brain mappings
In primates, there is a cortical magnification of certain had been included, the study was closed.
body parts (hand, mouth) in the postcentral gyrus (Kaas
et al. 1979; Nelson & Chen, 2008; Merzenich et al. 2014) Inclusion criteria
with a sequential, somatotopic organization of each finger
Data from successive brain mappings were prospectively
in narrow strips (Shoham & Grinvald, 2001). In hand
collected by the same team using the same protocol
area 3b, there is a rostral-to-caudal arrangement with the
throughout the 8 years of the study (June 2008–June
fingertip located rostrally (Paul et al. 1972; Kaas et al.
2016). Two main conditions of inclusion were defined:
1979; Merzenich et al. 2014). In contrast, in area 1, the
patients should have no initial sensory deficit and no brain
organization of the sensitivity of the phalanx is reversed,
lesion directly located in the postcentral gyrus. Patients were
with the fingertip located more caudally (Paul et al. 1972;
examined regarding their absence of initial sensory deficit
Kaas et al. 1979; Merzenich et al. 2014). Such reversal
using clinical tests (see Table 1 for preoperative testing
of the maps between 3b and 1 could be a functional
details). Before the operation, each patient underwent
criterion of 3b/1 limits (Sanchez-Panchuelo et al.
three tests: the tuning fork test; the warm/cold test; and
2012).
the two-points discriminative test on the cheek, the hand
Penfield and his co-workers (Penfield & Rasmussen,
(index finger) and the foot (hallux).
1950; Penfield & Jasper, 1954) mostly described the
According to Weinstein (1968), the mean threshold
functional anatomy of this area in humans, emphasizing
discriminative distance is around 4 mm in the index (both
the somatotopic organization of the hemibody. More
sexes), 7 mm in the cheek (both sexes), and 11 mm for
recently, activation studies have also found a somatotopic
males and 13 mm for females in the hallux, with no
organization of the somatosensory hemibody with more
differences between the left and right sides. Our criteria for
activation overlapping in areas 1 and 2 (Nelson & Chen,
exclusion were: patients with a discrimination threshold
2008) than in area 3 (Krause et al. 2001). Most studies
double the normal and/or any perturbed tuning fork or
underline the high functional (Sanchez-Panchuelo et al.
warm/cold test results on any of the three tested regions.
2012; Martuzzi et al. 2014; Kolasinski et al. 2016) and
anatomical (White et al. 1997) variability of primary
somatosensory maps. Pathology treated
However, since the seminal works of Penfield et al., no
Electrostimulation for brain mappings was performed
systematic mapping of the human somatosensory cortex to help with the removal of recently discovered brain
has been performed in a large number of subjects. The pre-
lesions. The mean time between the first clinical sign and
sent study was based on 50 patients with no somatosensory
operation was 38 days (range 3–150 days; SD 37 days).
deficit (‘intact’ patients). The aims of this prospective We found 15 WHO grade I and II gliomas, 25 WHO
electrostimulation study were to acquire accurate
grade III and IV gliomas, 4 arteriovenous malformations
coordinates of the functional areas of the somatosensory
or cavernomas, 5 metastases and 1 grade II meningioma.
homunculus in the standard MNI space and to study They were located in 29 right and 21 left hemispheres. The
the human hand and tongue cortical somatotopies in
mean age of patients was 49.5 years (range 29–73 years;
particular. Parameters of excitability of human primary
SD 14 years). Nineteen of the patients were women.
somatosensory areas in conscious individuals are also Thirty-seven patients had been recently treated with anti-
discussed.
epileptic drugs, always less than 3 months before the
operation. None of them had chronic intractable epilepsy.


C 2017 The Authors. The Journal of Physiology 
C 2017 The Physiological Society
J Physiol 000.00 Human somatosensory homunculus 3

Table 1. Results of pre-operative testing of the 50 patients

Tuning fork test Warm/cold test Two-point threshold discriminative test

Patients/sex Hand Cheek Foot Hand Cheek Foot Hand (index) Cheek Foot (hallux)

1/F 1 1 1 1 1 1 3 6 16
2/M 1 1 1 1 1 1 5 12 18
3/F 1 1 1 1 1 1 6 13 22
4/F 1 1 1 1 1 1 4 11 14
5/F 1 1 1 1 1 1 4 13 23
6/M 1 1 1 1 1 1 3 10 20
7/M 1 1 1 1 1 1 2 7 14
8/F 1 1 1 1 1 1 7 9 16
9/M 1 1 1 1 1 1 3 7 19
10/F 1 1 1 1 1 1 2 11 19
11/M 1 1 1 1 1 1 4 5 16
12/F 1 1 1 1 1 1 5 13 23
13/M 1 1 1 1 1 1 7 7 14
14/M 1 1 1 1 1 1 6 8 10
15/M 1 1 1 1 1 1 2 5 13
16/M 1 1 1 1 1 1 2 5 17
17/M 1 1 1 1 1 1 6 10 21
18/M 1 1 1 1 1 1 3 6 18
19/M 1 1 1 1 1 1 4 9 15
20/F 1 1 1 1 1 1 5 12 21
21/M 1 1 1 1 1 1 3 8 17
22/F 1 1 1 1 1 1 5 11 20
23/M 1 1 1 1 1 1 6 12 19
24/F 1 1 1 1 1 1 2 8 18
25/M 1 1 1 1 1 1 4 7 17
26/M 1 1 1 1 1 1 2 5 8
27/M 1 1 1 1 1 1 3 11 18
28/M 1 1 1 1 1 1 4 10 21
29/F 1 1 1 1 1 1 2 10 24
30/F 1 1 1 1 1 1 2 8 21
31/M 1 1 1 1 1 1 2 8 21
32/F 1 1 1 1 1 1 4 11 16
33/F 1 1 1 1 1 1 3 13 20
34/M 1 1 1 1 1 1 5 12 21
35/M 1 1 1 1 1 1 3 10 19
36/M 1 1 1 1 1 1 2 7 17
37/M 1 1 1 1 1 1 3 6 15
38/M 1 1 1 1 1 1 4 8 16
39/F 1 1 1 1 1 1 2 9 18
40/M 1 1 1 1 1 1 2 11 12
41/M 1 1 1 1 1 1 3 9 13
42/F 1 1 1 1 1 1 2 9 10
43/M 1 1 1 1 1 1 2 12 11
44/F 1 1 1 1 1 1 4 8 14
45/M 1 1 1 1 1 1 4 9 15
46/M 1 1 1 1 1 1 2 6 9
47/M 1 1 1 1 1 1 2 6 10
48/F 1 1 1 1 1 1 5 7 8
49/F 1 1 1 1 1 1 4 4 10
50/M 1 1 1 1 1 1 5 4 14
Tuning fork test: Do you feel the tuning fork? Yes: 1; no: 0. Warm/cold test: Do you feel warm/cold? Yes: 1; no: 0. Two-point threshold
discriminative test: None of the patients of our study had a discriminative deficit in the tested zones and none failed the tuning fork
or warm/cold tests in any of the 3 different regions tested.Discrimination test thresholds:Whole group for index: Mean discriminative
test, 3.58 mm. Standard deviation: 1.47.Whole group for cheek: Mean discriminative test, 8.78. Standard deviation: 2.61.Whole group
for hallux: Mean discriminative test, 16.42. Standard deviation: 4.16.


C 2017 The Authors. The Journal of Physiology 
C 2017 The Physiological Society
4 F.-E. Roux and others J Physiol 000.00

Anaesthetic protocol for awake craniotomy The patient’s level of alertness in the absence of stimulation
was regularly evaluated throughout the testing, as further
Our awake brain mapping protocol was based on
assurance that changes during stimulation were not
20 years’ experience (Roux et al. 2016). Anaesthetic drugs
random events. Non-reproducible interferences (positive
can, in theory, interfere with stimulation thresholds.
responses to electrostimulation) were not included in this
Our objective during brain mapping was to avoid any
study. The reproducibility criterion was 3/3 (i.e. 3 inter-
anaesthetic drugs. One hour before admission to the
ferences to validate a cortical site as ‘positive’). When a
operating room, a patch containing a eutectic mixture
site had a reproducibility of 2/3 we stimulated it at least
of prilocaine (2.5 mg g−1 ) and lidocaine (2.5 mg g−1 )
one more time: reproducibility criteria of 3/4 (or 4/5) were
(EMLA) was applied in the supraorbital and auricu-
validated but not 2/4 or 3/5. At least 3 trials were performed
lotemporal regions. Lidocaine (1%) with adrenaline
on positive sites.
(epinephrine; 1:100,000) was infiltrated to block the
To evaluate the current amplitude for somatosensory
supraorbital, auriculotemporal, and occipital nerves.
mappings in postcentral gyri, the current amplitude was
Additionally, the Mayfield head holder (Ohio Medical,
started at 1 mA and progressively increased by 0.5 mA
Cincinnati, OH, USA) pin site and the surgical skin
steps. If allowed by electrocorticography results, the
incision line were infiltrated. Sedation with spontaneous
intensity of stimulation was raised to as high as 10 mA
respiration was provided by continuous infusion of
if necessary to obtain a response. If patients felt any
Propofol (1–3 mg kg h−1 ). Fentanyl (1–3 µg kg h−1 ) or
unpleasant sensation or pain at any time, or any sensation
Remifentanil (0.01–0.25 µg kg h−1 ) was used for analgesia.
that could evoke seizures, the stimulation was stopped and
The depth of procedural sedation was adjusted to keep the
the intensity reduced.
patient’s vital signs stable. Propofol infusion was stopped
Because our ability to test the human brain was res-
during the dural opening (around 10 min before brain
trained by clinical requirements, the possibility of deli-
mapping) and the patient was fully awakened. Once the
vering stimulations was limited. We tested only what was
cortical mapping procedure was completed, patients were
really useful for the treatment of the patients. Stimulation
put back to sleep using the same protocol for the rest of
around and between detected positive areas may not have
the operation.
been tested because such testing was not clinically relevant
for the patient. This legitimate constraint limited the
ability to find complete somatotopies of the part of the
Cortical procedures
body tested in individual patients.
A neuronavigational system was used to guide tumour The presupposition of this study in terms of localization
removal. Anatomical structures (gyri and sulci) were was that, as described in other human anatomical,
identified according to this neuronavigational data and activation, or receptor-binding studies (Geyer et al. 1997;
the visual identification of the shape of gyri and sulci. Nelson & Chen, 2008; White et al. 1997), BA1 was defined
The cortex was directly stimulated before any surgical as the crown of the postcentral gyrus from the central
approach using the bipolar electrode of the Nimbus sulcus anteriorly to the postcentral sulcus. These divisions
cortical stimulator (1 mm electrodes; Innopsys, Toulouse, are similar to those used for primate studies (Mountcastle,
France) with biphasic square wave pulses of 1 ms duration 2005). Because of their localizations deep in the sulci,
and 50 Hz trains. The maximum train duration of BA3a, BA3b, and BA2 were not accessible to electro-
each stimulation was 5 s. The afterdischarge threshold stimulation. Thus, the current functional exploration
was determined by electrocorticography using a strip focused on BA1.
electrode. Although this point remains controversial, the In the 50 patients, 647 sites were stimulated overall on
afterdischarge threshold seemed to change little from site the postcentral gyrus and 1025 in gyri nearby (Table 2).
to site during one short mapping session (Ojemann, When a functional site was found, it was marked by a
1993). The level of electrostimulation was always kept sterile ticket of 0.25 mm2 , identified in neuronavigation
1 mA below that expected to cause electrical diffusion and before moving on to the next test site. It was decided that
afterdischarges so as to ensure that the stimulated area the minimal spatial resolution of our electrostimulation
remained accurately localized on the area of cortex under technique corresponded to the size of the bipolar electrode
study. If any afterdischarge (or epilepsy) was detected, separated by 3 mm. Once a positive functional area was
the protocol was adapted to the afterdischarge threshold detected, the cortical areas located 3 mm from it were also
found. tested.
Only clear brain mapping data were included in
this study. Patients’ feedback was fundamental: we con-
Postsurgical and statistical analysis
sidered that a response was obtained when the
patients acknowledged feeling something. Patients were Each patient had her/his positive stimulations positioned
encouraged to report their sensations during stimulation. on the left or right 3D cortical surface reconstructions of


C 2017 The Authors. The Journal of Physiology 
C 2017 The Physiological Society
J Physiol 000.00 Human somatosensory homunculus 5

Table 2. Number and localization of cortical sites stimulated during stimulation. Pain was felt only once by one patient;
in patients another also had a bad taste in the mouth and another
reported burning in the hand during some stimulations.
Left and right hemispheric Number of stimulations
No other sensations (from the proprioception domain for
regions stimulated (n = 1672)
instance) were detected. Strictly no other isolated, positive
Postcentral gyri 647 somatosensory sensation was found among the 1025
Precentral gyri 208 stimulations performed outside the postcentral gyrus.
Supramarginal gyri 350
Upper parietal 189
Superior temporal gyri 153
The medial-to-lateral sequence of the hand
Middle temporal gyri 72 somatotopy
Inferior frontal gyri 53
As expected, the ventral parts of the fingers were
somatotopically organized from the medial-to-lateral
aspects of postcentral gyrus and were symmetrically
similar in the right and left hemispheres (Fig. 2A–C).
one of the individual brains (case 12) constituting the PALS Finger responses were found in 34 patients and the
atlas (Van Essen et al. 2005) provided in the Caret software medial-to-lateral somatotopy was respected with no
(Van Essen et al. 2001) and normalized in the MNI aberrant localization. Nevertheless, some rare individual
space. We obtained normalized coordinates of stimulation variations were noted: two stimulations in two different
site locations that were per-operatively visualized and patients elicited tingling not in individual fingers but in
positioned on original 3D images provided by the neuro- all fingers. They were located in the main postcentral
navigation software (Brain Lab). For each positive site, somatotopic representation of the digits. Only one patient
MNI space coordinates (X, Y, Z) were obtained and stored reported sensations in the dorsal part of her fingers when
year by year in an Excel database, with a detailed account one cortical area was stimulated. One stimulation elicited
of the type of response obtained. tingling in the little finger and the medial part of the nearby
ring finger. More areas were found for the thumb (41
areas) and 2nd finger (31 areas), than for the 3rd finger
Results (25 areas), the 4th finger (19 areas) and the little finger (21
Parameters of stimulation of the somatosensory areas). All MNI coordinates of finger positive points are
available in Table 3A. We found a certain dispersion of the
primary areas in the conscious human
points between the right and left hemispheres. Dispersions
Over 647 sites stimulated in BA1, we found 258 positive were computed by calculating the Euclidean distance (in
areas (40%). Stimulation of somatosensory primary areas mm) between each stimulation point and the barycentre
had several characteristics (positive intensity threshold; of the corresponding finger. A Student’s paired t test across
type of outbreak; accuracy; type of patients’ feelings). The clusters indicated significant differences both in terms of
mean general positive intensity threshold was 2.67 mA mean distance (t = 3.2, P < 0.05) and standard deviations
(from 0.5 to 7 mA) with a limited standard deviation (t = 3.7, P < 0.01) between the left and right hemispheres.
(SD 0.72 mA). Specific mean finger (2.39 mA) and This was probably due to the template we used, which was
tongue (2.60 mA) positive thresholds were not statistically not strictly symmetric. However, there was no significant
different, unlike the intensities required to elicit sensations difference in the distances from the barycentres of each
in the other body parts (legs, shoulder, trunk, arms), which finger between the two hemispheres (t = 0.4, P = 0.68).
were significantly higher (mean = 3.5 mA; P < 0.05 with
the Wilcoxon test). No significant difference for positive The rostral-to-caudal sequence of the hand
thresholds was found in patients with or without anti-
somatotopy
epileptic drugs (P = 0.80; Wilcoxon test). No correlation
was detected between the patient’s age group (comparing In addition to this medial-to-lateral sequence, 10 patients
patients over 60 years and younger patients) and the level felt either the tip or the base of at least one finger during
of the positive thresholds (P = 0.4576; Spearman test). electrostimulation. Furthermore, four patients displayed
The somatosensory positive areas were located in very both tip and base finger phalange representation, all of
small patches of cortex of around 3 mm by 3 mm (Fig. 1). them with a caudal discrete representation of the finger
Positive stimulations were detected with a clear-cut tip. Thus, group analysis of these 14 patients who displayed
‘threshold’ effect (for instance, at 1.5 mA, the subject felt base and/or tip of the finger sensations during stimulation
nothing and then, at 2 mA, the subject felt something). showed a clear rostral-to-caudal finger somatotopy with
Tingling (discriminative touch domain) was the main the tips of the fingers located more caudally than the base,
and almost exclusive sensation reported by all patients as shown in Fig. 2D and E.


C 2017 The Authors. The Journal of Physiology 
C 2017 The Physiological Society
6 F.-E. Roux and others J Physiol 000.00

The tongue somatotopy est (thumb, index, middle finger, middle part of the
tongue, and lips), calculated from MNI coordinates, were
Fourteen patients displayed a medial-to-lateral hemi-
all less than 5 mm.
spheric tongue somatotopy with the tip of the tongue
located more medially than the middle of the tongue
or its base (Fig. 3). Patients felt tingling exclusively on
the contralateral space of the tongue. Most sensations Discussion
concerned the middle of the tongue (31 of 44 sites). Tip
sensations were rather rare (6 sites), as were sensations Since the seminal works of Penfield and collaborators
of the base of the tongue during stimulation (7 sites). (Penfield & Rasmussen, 1950; Penfield & Jasper, 1954), no
Overall, the medial-to-lateral space devoted to the tongue other work has studied the human somatosensory cortex
in BA1 was quite large, around 2.5 cm. Although with so many subjects, whatever the mapping technique.
a rostral-to-caudal somatotopy (for instance different The analysis of MNI coordinates demonstrated that BA1
cortical representation for the lateral and the central parts of the somatosensory cortex in humans was organized
of the tongue) may exist for the human tongue, no such somatotopically from the medial to lateral and anterior
organization was detected. All MNI coordinates of tongue to posterior parts of the postcentral gyrus. The contra-
positive points are available in Table 3B. lateral cutaneous areas of the body are represented in small
patches of cortex and, although almost all cutaneous parts
of the contralateral body were found at least once in this
The medial-to-lateral sequence of the contralateral series, two main zones were detected: the hand and the
face/tongue. Other body zones were found much more
body
rarely. The hand and the face/mouth areas occupied parts
Figure 4 shows the somatotopic sequence of the of the cortex that were proportionally much larger than
human body with MNI coordinates. Various contralateral their cutaneous surface area, because of their much more
representations of the skin from the feet, knee, abdomen, marked discriminative properties and small receptive
thorax, eyebrows, gums or jaws were detected in BA1, fields. The relatively high number of positive responses
among other sensations. All MNI coordinates of body found for hand and face/tongue is an indirect sign that
positive points are available in Table 3C. As shown in the human brain multiplies the cortical representations of
Fig. 5 and in Table 4, the variations in localization were small receptive fields to increase its discriminative abilities
limited. The standard deviations of main regions of inter- for these two body regions. More precisely, in the hand

Figure 1. Example of electrostimulation

mapping for somatosensory BA1 cortex
A, 38-year-old patient with a right diffuse
astrocytoma IDH mutant (WHO 2016) with
B no sensorimotor deficit. B, 3D brain
localization of the positive stimulations. In
the postcentral gyrus (BA1),
electrostimulation showed a clear finger
somatotopy, in small (3 mm wide), sharply
delimited cortical areas from the little (MNI
coordinates: X = 39.4, Y = -31.2,
C Z = 64.2), ring (42.9, −29.0, 60.4), middle
(44.7, −24.8, 59.5), and index (−44.3,
−25.8, 59.0) fingers to the thumb (46.0,
−25.7, 50.7). We observed that the
displacement of the bipolar electrode on
the cortex located adjacent to a positive
area often induced another interference. C,
accuracy of the technique with 3 mm
distance between the stimulating
electrodes. [Colour figure can be viewed at
wileyonlinelibrary.com]


C 2017 The Authors. The Journal of Physiology 
C 2017 The Physiological Society
J Physiol 000.00 Human somatosensory homunculus 7

A B
−15
60 Left

Anterior
50 −20

40

Y (mm)
−25
30
−30
20
−35
10
0 −40
−10 −55 −50 −45 −40 −35 −30 −25
Y (mm)

−20 x (mm)

−30 C
−40 −15
−50
−20
−60

Y (mm)
−25
−70
Posterior

−80 −30
−90 −35
−100 Left Right Right
−40
−60−50−40−30−20−10 0 10 20 30 40 50 60 25 30 35 40 45 50 55
x (mm) x (mm)

D E
−15 Anterior Anterior
Left 1
bank
bank
−20
0.9
−25
Relative distance: Dposterior / (Dposterior+Danterior)
Y (mm)

0.8
−30
0.7
−35
Posterior
−40 bank 0.6

−55 −50 −45 −40 −35 −30 −25

0.5

−15 0.4
−20
0.3
Y (mm)

−25
0.2
−30
0.1
−35
Right Posterior
−40 0
bank
25 30 35 40 45 50 55
x (mm)

Figure 2. The medial-to-lateral and rostral-to-caudal sequences of the hand somatotopy

A, distribution over the two hemispheres (left and right templates) of the medial-to-lateral somatosensory
representation of each finger. B and C, representation of the barycentre of each finger (coloured dots) in right
and left hemispheres with amplitude bars representing the standard variations of each localization (coloured
lines). Average right and left MNI barycentres (X, Y, Z) of the thumb (46.2, −22.5, 56.7), index (44, −22.8,


C 2017 The Authors. The Journal of Physiology 
C 2017 The Physiological Society
8 F.-E. Roux and others J Physiol 000.00

59.9), middle finger (40.6, −28.2, 62.1), ring (37.7, −29.2, 64.8) and little finger (35.4, −30, 66.3). Overall, the
medial-to-lateral space (X MNI coordinates) devoted to the fingers in BA1 was around 1.5 cm. D, right and left
hemispheric distribution of the MNI coordinates of the somatosensory base of the fingers (yellow); whole finger
sensation (grey); and tips of the fingers (light blue). Individual analysis showed the rostral-to-caudal somatosensory
cortical representation of the ventral part of the skin of the fingers. This rostral-to caudal somatotopy was detected
at least once in all 5 fingers. E, group analysis confirmed this somatotopic representation with average X, Y, Z MNI
coordinates for the group of fingertip points located more posteriorly (X = 41.4; Y = −31.1, Z = 61.1) than the
whole finger sensations (39.7, −27.3, 62.9) and base of the finger sensations (43.6, −26.1, 60.1). [Colour figure
can be viewed at wileyonlinelibrary.com]

area, magnifications of the ventral part of the thumb varied little (standard deviations of the barycentre of main
and index were particularly evident, as described in other cortical representations were less than 5 mm).
studies (Penfield & Rasmussen, 1950; Martuzzi et al. 2014; Although the medial-to-lateral organization was
Sanchez-Panchuelo et al. 2012). described long ago, the anterior-to-posterior organization
Finally, activation (Sanchez-Panchuelo et al. 2012; of the postcentral gyrus in the human is less well known
Kolasinski et al. 2016) or anatomical (White et al. 1997) (Nelson & Chen, 2008). This study showed that the
studies argue for the variability of this somatosensory finger topography is organized with the tip of the fingers
functional organization. The main question concerns what located more posteriorly than their base, in both group
is considered as ‘variability’ in a functional organization. and single subject analysis. This finding matches previous
Our results showed that no aberrant somatotopic studies in primates that suggested such a rostral-to-caudal
organization was detected and the localization of each finger organization in BA1 (Darian-Smith et al. 1984;
skin cortical representation within the postcentral gyrus Mountcastle 2005). It is also in line with previous

70 Left 70 Right

60 60

50 50
Z (mm)
Z (mm)

40 40
Tongue Tongue
2

1 Tip 2 3 Tip
30 30
4
1 Middle 2 3 Middle

1 3 4
20 Base 20 Base
4

L R L R
10 10

10 0 −10 −20 −30 −30 −20 −10 0 10

Y (mm) Y (mm)

Figure 3. The somatotopic sequence of the tongue

The contralateral somatotopic representations displayed the following somatotopy: base of the tongue (red
dots) close to the Sylvian fissure, and middle (pink dots) and tip of the tongue (yellow dots) more medial. The
representation of the tongue occupied a large portion of the postcentral gyrus (2.5 cm for the mouth). This figure
is composed of isolated points found punctually in some patients and of multiple points found in four individual
patients (marked 1, 2, 3 and 4 in square boxes). Although a certain degree of variation may exist within this part
of the postcentral gyri in the exact location of the tip, middle, and base of the tongue somatosensory cortical
localizations, individual somatotopy respects the tip/medial–base/lateral tongue somatotopy. [Colour figure can be
viewed at wileyonlinelibrary.com]


C 2017 The Authors. The Journal of Physiology 
C 2017 The Physiological Society
J Physiol 000.00 Human somatosensory homunculus 9

Table 3. MNI coordinates, barycentres (bold) and standard deviations (italics) of main region of interest for all 258 positive points
(specific cutaneous sensation reported by patients)

A. Fingers (1)

Thumb Index finger Middle finger

X Y Z Pat. X Y Z Pat. X Y Z Pat.

−44 −25 59 1 −45 −26 58 4 −46 −26 58 1

−47 −28 56 1 −41 −29 62 8 −44 −31 57 1
45 −22 59 2 40 −29 64 9 37 −30 66 9
−48 −25 55 4 −48 −28 54 10 −42 −29 60 10
−49 −26 53 7 43 −26 61 11 41 −28 63 11
−47 −26 57 8 41 −29 64 12 37 −32 66 12
−43 −27 59 8 46 −23 58 13 44 −26 60 13
45 −21 59 9 −45 −24 58 14 −42 −26 60 14
42 −27 63 9 −47 −28 56 14 −47 −26 57 15
−51 −25 51 10 44 −22 60 16 42 −26 62 16
44 −23 60 11 42 −28 62 17 43 −24 62 17
46 −22 58 12 42 −28 62 19 −41 −27 62 21
47 −19 56 13 46 −23 58 20 −40 −26 62 22
−48 −23 54 14 −46 −26 58 21 −41 −26 63 23
−50 −23 51 14 −48 −30 56 21 −43 −31 58 26
−47 −25 56 15 −43 −24 60 22 45 −25 59 27
44 −19 58 16 −45 −25 58 23 −41 −26 62 28
47 −19 55 16 43 −27 61 24 −42 −32 60 28
45 −21 59 17 −47 −29 57 26 40 −30 64 34
44 −25 61 17 44 −26 59 27 42 −31 61 34
46 −20 57 20 −46 −18 54 28 38 −29 65 40
−48 −23 54 21 −48 −30 55 28 38 −32 65 40
−47 −22 55 22 43 −24 61 34 34 −34 66 41
−47 −24 56 22 42 −27 62 34 33 −32 68 41
−51 −24 51 23 45 −26 59 34 36 −33 65 50
−49 −23 53 23 40 −28 63 40
45 −22 59 24 41 −31 62 40
−49 −24 53 26 34 −34 66 41
46 −26 51 27 36 −31 67 41
−47 −24 56 28 −44 −25 59 44
48 −20 55 30 43 −27 61 48
−47 −24 55 32
44 −16 60 33
47 −19 57 34
45 −22 59 34
44 −21 59 40
−45 −24 58 44
−48 −22 53 44
−55 −22 46 46
45 −21 59 48
44 −25 61 48
46.6 −22.8 56.2 43.4 −26.8 59.9 40.8 −28.6 62.0
2.49 2.58 3.44 3.25 3.14 3.23 3.51 2.93 3.06
(Continued)


C 2017 The Authors. The Journal of Physiology 
C 2017 The Physiological Society
10 F.-E. Roux and others J Physiol 000.00

Table 3. Continued

A. Fingers (2)

Ring finger Little finger

X Y Z Pat. X Y Z Pat.

−41 −26 63 1 −39 −29 65 1

−41 −27 63 4 −37 −30 66 4
30 −32 68 9 24 −32 71 9
37 −31 66 11 34 −31 67 11
29 −33 69 12 25 −34 71 12
43 −28 62 13 40 −31 63 13
−40 −27 63 14 −37 −29 66 14
41 −28 63 16 −42 −28 62 15
42 −29 62 17 38 −30 66 16
−38 −27 64 22 38 −31 65 17
−37 −27 65 23 −35 −28 65 22
35 −32 67 24 −35 −29 66 23
−41 −30 62 26 −39 −32 65 26
43 −29 60 27 39 −31 64 27
−38 −28 65 28 −36 −31 66 28
34 −32 67 34 39 −30 64 30
35 −30 67 40 33 −31 67 40
33 −33 67 40 31 −34 69 40
34 −34 66 41 34 −34 66 41
31 −33 69 41
32 −34 68 50
37.5 −29.7 64.8 35.2 −30.9 66.3
4.23 2.47 2.47 4.65 1.93 2.34

B. Tongue

Base Middle Tip

X Y Z Pat. X Y Z Pat. X Y Z Pat.

60 −10 25 3 61 −10 29 3 61 −11 34 3

−62 −13 18 18 61 −12 34 6 −57 −11 38 15
62 −14 23 31 −58 −9 34 15 61 −9 30 36
61 −11 25 38 −60 −8 31 18 61 −10 32 38
61 −11 28 43 −60 −10 26 18 58 −13 42 43
−62 −11 23 46 −61 −11 23 18 −57 −12 40 46
62 −8 21 49 −61 −11 31 18
−62 −13 28 18
−63 −14 24 18
59 −9 26 31
62 −13 28 31
62 −14 32 31
61 −9 33 31
61 −10 34 36
59 −12 38 36
60 −10 28 37
61 −11 31 37
61 −10 29 38
−61 −10 32 42
−58 −9 34 42
−57 −12 40 42
61 −10 32 43
(Continued)


C 2017 The Authors. The Journal of Physiology 
C 2017 The Physiological Society
J Physiol 000.00 Human somatosensory homunculus 11

Table 3. Continued

B. Tongue

Base Middle Tip

X Y Z Pat. X Y Z Pat. X Y Z Pat.

58 −11 37 43
62 −14 34 43
61 −15 38 43
−62 −14 19 45
−62 −12 24 45
−63 −16 23 45
−61 −13 31 46
−61 −10 28 46
−60 −9 32 46
−58 −9 35 46
−62 −12 24 47
60 −10 25 49
61 −10 33 49
62 −15 28 49
62 −15 22 49
61.4 −11.1 23.3 60.7 −11.4 30.0 59.2 −11.0 36.0
0.78 1.95 1.41 1.47 2.12 4.97 2.04 1.41 4.73

C. Body (standard deviation for the palm and lips only)

Feet Leg Knee

X Y Z Pat. X Y Z Pat. X Y Z Pat.

−4 −41 64 1 −5 −44 67 1 8 −41 78 50

3 −42 76 50
4 −41 64 4 −43 71.5 8 −41 78

Thigh Hip Abdomen

X Y Z Pat. X Y Z Pat. X Y Z Pat.

−9 −42 73 1 −19 −37 76 14 −22 −37 72 26

−17 −38 76 26
11 −41 77 50
12.3 −40.3 75.3 19 −37 76 22 −37 72

Thorax Shoulder Elbow

X Y Z Pat. X Y Z Pat. X Y Z Pat.

−13 −41 75 1 −33 −35 69 5 26 −32 70 13

16 −40 75 9 −31 −36 70 26 −31 −35 70 14
16 −40 75 50 −25 −35 71 32 19 −33 72 19
−27 −35 71 28
15 −40.3 75 29.7 −35.3 70 25.7 −33.7 70.7

Forearm Wrist Palm

X Y Z Pat. X Y Z Pat. X Y Z Pat.

36 −32 66 13 −32 −34 69 1 43 −26 62 2

28 −32 69 16 −35 −34 68 14 44 −21 59 12
27 −34 70 24 25 −32 71 19 43 −24 62 13
(Continued)


C 2017 The Authors. The Journal of Physiology 
C 2017 The Physiological Society
12 F.-E. Roux and others J Physiol 000.00

Table 3. Continued

Forearm Wrist Palm

X Y Z Pat. X Y Z Pat. X Y Z Pat.

30 −34 69 25 −35 −34 68 28 42 −25 62 16

25 −35 70 41 27 −34 70 29 −39 −28 65 21
33 −32 68 30 −48 −27 54 22
−29 −36 70 32 −42 −25 61 23
44 −23 60 25
44 −23 60 29
44 −28 60 40
37 −29 66 41
38 −30 65 50
29.2 −33.4 68.8 30.9 −33.7 69.1 42.3 −25.7 61.3
3.05 2.73 3.22

Eyebrows Eyes Nose

X Y Z Pat. X Y Z Pat. X Y Z Pat.

−53 −23 48 10 −56 −21 46 8 −56 −19 45 10

53 −23 48 56 −21 46 56 −19 45

Face Lips Teeth

X Y Z Pat. X Y Z Pat. X Y Z Pat.

−55 −15 44 22 55 −13 47 13 62 −8 21 43

−57 −13 42 15 60 −10 28 43
52 −13 51 34
60 −14 40 37
59 −13 40 38
57 −13 44 38
60 −13 38 39
57 −12 44 40
59 −13 40 40
57 −13 44 42
−62 −10 28 45
57 −17 43 46
−56 −16 43 46
−57 −13 42 46
55 −15 44 57.5 −13.3 41.9 61 −9 24.5
2.44 1.63 4.88

Gums Jaw Pharynx

X Y Z Pat. X Y Z Pat. X Y Z Pat.

58 −10 37 6 59 −14 42 13 −62 −14 14 18

61 −14 38 13
61 −12 34 37
61 −12 36 37
58 −10 37 60.5 −13 37.5 62 −14 14

Larynx

X Y Z Pat.

62 −7 19 35
−62 −13 18 46
62 −10 18.5
Pat. = patients.


C 2017 The Authors. The Journal of Physiology 
C 2017 The Physiological Society
J Physiol 000.00 Human somatosensory homunculus 13

findings of some activation studies performed in humans Human somatosensory positive thresholds could be
(Blankenburg et al. 2003; Sanchez-Panchuelo et al. 2012). modified by other confounding factors (Haglund et al.
The coordinate of the caudal discrete representations of 1993) such as current frequencies and monopolar versus
fingertips in area 1 found in group analysis (X = −47; bipolar stimulation devices, the type of anaesthesia
Y = −31; Z = 53) by Blankenburg et al. (2003) are used, stimulations during the refractory period, patients’
extremely close to our coordinates (X = 41.4; Y = −31.1, sensory skills or the presence of a sensory deficit. These
Z = 61.1). The human cortical representation of the possibilities need to be investigated by a subsequent study.
tongue occupies a territory at least as large as that of Many patches of cortex were observed to be unresponsive
the hand between, medially, the lips and, more laterally, the to stimulation. In our opinion, the reason may have
pharynx, but its organization is radically different for the been that the neurons under the electrostimulation
tongue, with the tip of the tongue localized more medially probe did not respond in a way visible to us during the
than its base. operation.
Lastly, in somatosensory mapping, the responses we
obtained depended only on the ability of the patient
Human somatosensory cortex positive thresholds to be conscious of his/her sensation and to report it.
Nothing is known about the possible diffusion of this
In those patients with intact postcentral gyrus, soma- stimulation in deep brain structures. The stimulation
tosensory responses were obtained with mean thresholds could be located on a functional area without the sensation
around 2.5 mA in the tongue and hand areas and with reaching the patient’s consciousness, a phenomenon
slight individual variations. The positive threshold levels described in language mappings (Roux et al. 2015).
needed to obtain thoracic or lower limb sensations were Clinical information (and further scientific data) could
higher, and levels were different for other parts of the thus have been missed.
human brain (Roux et al. 2016). We found that the age of
the patients and the presence of antiepileptic drugs had no
effect on these thresholds. Nevertheless, other factors that
Differences with the previous works of Penfield et al.
we did not test could also intervene. For instance, patients
who had had previous chemotherapy could have different Although our study matches many findings by Penfield
thresholds. and co-workers, four main differences can be noted.

A Left Right B Feet

60 Leg
−20
50 Knee
40 −30 Tigh
Y (mm)

Hip
30
−40 Abdomen
20 Thorax
10 −50
Shoulder
0 Elbow
−10 −50 −40 −30 −20 −10 0 0 10 20 30 40 50 Forearm
Y (mm)

X (mm) X (mm) Wrist

−20
−30
Eyebrows
−40 60
Eye
−50 Nose
50
−60 Face
Z (mm)

40 Lips
−70
−80 30 Teeth
Gums
−90 20 Jaw
−100 Pharynx
10
Larynx
−60−50−40−30−20−10 0 10 20 30 40 50 60 10 0 −10 −20 −30 −40 −40 −30 −20 −10 0 10
Tongue
X (mm) Y (mm) Y (mm)

Figure 4. The somatotopic sequence of the body

A, global somatotopic organization of the body (flesh colour), hand (mauve), and head (green). Dotted boxes:
regions of interest for body and face somatotopies. B, top: somatotopic representation of the limbs with their
medial-to-lateral somatotopic sequence. B, bottom: somatotopic representation of the face (excluding tongue
represented in Fig. 3). The medial-to-lateral somatotopic sequence of the eyebrows, eyes, cheeks, both lips, teeth,
gums, jaws, pharynx, and larynx is shown. [Colour figure can be viewed at wileyonlinelibrary.com]


C 2017 The Authors. The Journal of Physiology 
C 2017 The Physiological Society
14 F.-E. Roux and others J Physiol 000.00

Firstly, they found that 25% of the somatosensory Table 4. Standard deviations (in mm) of MNI coordinates
responses were outside the postcentral gyrus, in the pre- of positive points for main regions of interest calculated
central gyrus or parietal lobe (Penfield & Rasmussen,
Region of interest X Y Z
1950; Penfield & Jasper, 1954). They also described a small
‘second sensory’ area in the superficial bank of the Sylvian Thumb (41 points) 2.49 2.58 3.44
fissure of the precentral gyrus (Penfield & Roberts, 1959). Index (31 points) 3.25 3.14 3.23
We detected no somatosensory responses outside the post- Middle finger (25 points) 3.51 3.14 3.06
central gyrus. A second difference is that we recorded no Middle part of tongue (37 points) 1.47 2.12 4.97
bilateral responses for stimulation in the face and tongue Lips (14 points) 2.44 1.63 4.88
regions, whereas 13% (27 out of 202) of the responses in
Penfield and collaborators’ works were bilateral (Penfield
& Rasmussen, 1950). One hypothesis is that paradoxical
responses may not have been uncommon in the epileptic
patients treated by Penfield, particularly in those with Penfield & Rasmussen (1950) did not identify in the
organic lesions and chronic epilepsy (Urasaki et al. 1994). cortex ‘because of its small size’. Seldom studied, the
Some of Penfield’s patients had huge brain lesions around somatotopic representation of the dorsal surface of
the central gyri and thus possible cerebral reorganizations. the digits could be comparable to that of the ventral
The findings of somatosensory responses outside the post- surface in human area 3b (Druschky et al. 2002). This
central gyri may not have been uncommon. finding is perhaps a question of opportunity. During
Thirdly, Penfield’s team also described some sensations their career, Penfield and colleagues used different types
produced by stimulation, such as a sense of movement of stimulators, the frequency of which was not always
and, more rarely, a desire to move (Penfield & Rasmussen, controlled (Penfield and Roberts, 1959). It is possible
1950; Penfield & Jasper, 1954), that were never reported that the accuracy of the stimulator we used, in terms of
by our patients. localization (with 3 mm wide electrodes) and stability
Fourthly, we reported rare sensations on the dorsal of the electric frequency, allowed us to detect such rare
part of the hand or fingers, a cutaneous territory sensations.
Middle finger
Index finger
Ring finger

Eyebrows
Abdomen

Shoulder

Forearm

Pharynx
Tongue
Thumb
Thorax

Larynx
Elbow

Gums
Pinky

Teeth
Knee

Nose
Wrist
Palm

Face
Feet

Tigh

Lips

Jaw
Eye
Leg

Left Right
Hip

80 0 0 0 0 80

70
0.2 0.2 70

60
60
0.4 0.4
50
50
40 0.6 0.6
40
30
0.8 0.8
30
20
1 1 1
1 20
10
−10 −20 −30 −40
10
−40 −30 −20 −10

Figure 5. Anatomical relative variations of the human somatosensory cortex

Representation of the relative variations of the human somatosensory cortex on a scale from 0 (medial part) to 1
(lateral part of the postcentral gyrus). [Colour figure can be viewed at wileyonlinelibrary.com]


C 2017 The Authors. The Journal of Physiology 
C 2017 The Physiological Society
J Physiol 000.00 Human somatosensory homunculus 15

References Mountcastle VB (2005). The Sensory Hand: Neural Mechanisms

of Somatic Sensation. Harvard University Press, Cambridge,
Blankenburg F, Meyer R, Schwiemann J & Villringer A (2003). MA, USA.
Evidence for a rostral-to-caudal somatotopic organization in Nelson AJ & Chen R (2008). Digit somatotopy within cortical
human primary somatosensory cortex with mirror-reversal areas of the postcentral gyrus in humans. Cereb Cortex 18,
in areas 3b and 1. Cereb Cortex 13, 987–993. 2341–2351.
Brodmann K (1909). Vergleichende Lokalisationslehre der Ojemann G (1993). Functional mapping of cortical language
Grosshirnrinde. Johann Bart, Leipzig, Germany. areas in adults. In Electrical and Magnetic Stimulation of the
Darian-Smith I, Goodwin A, Sugitani M & Heywood J (1984). Brain and Spinal Cord, eds Devinsky O & Beric A. Raven
The tangible features of textured surfaces: their Press, New York.
representation in the monkey’s somatosensory cortex. In Paul RL, Merzenich M & Goodman H (1972). Representation
Dynamic Aspects of Neocortical Function, eds Edelman GM, of slowly and rapidly adapting cutaneous mechanoreceptors
Gall WE & Cowan WM. Wiley & Sons, New York. of the hand in Brodmann’s areas 3 and 1 of Macaca mulatta.
Druschky K, Kaltenhäuser M, Hummel C, Druschky A, Pauli E, Brain Res 36,229–249.
Huk WJ, Stefan H & Neundörfer B (2002). Somatotopic Penfield W & Jasper H (1954). Epilepsy and the Functional
organization of the ventral and dorsal finger surface Anatomy of the Human Brain. Little, Bown and Company,
representations in human primary sensory cortex Boston.
evaluated by magnetoencephalography. Neuroimage 15, Penfield W & Rasmussen T (1950). The Cerebral Cortex of Man.
182–189. Hafner Publishing company, New York.
Gardner EP (1988). Somatosensory cortical mechanisms of Penfield W & Roberts L (1959). Speech and Brain Mechanisms.
feature detection in tactile and kinesthetic discrimination. Princeton University Press, Princeton, NJ, USA.
Can J Physiol Pharmacol 66, 439-454. Powell TP & Mountcastle VB (1959). The cytoarchitecture of
Geyer S, Schleider A & Zilles K (1997). The somatosensory the postcentral gyrus of the monkey Macaca mulatta. Bull
cortex of human: cytoarchitecture and regional distributions Johns Hopkins Hosp 105, 108–131.
of receptor-binding sites. Neuroimage 6, 27–45. Roux FE, Durand JB, Djidjeli I, Moyse E & Giussani C (2016).
Haglund MM, Ojemann GA & Blasdel GG (1993). Optical Variability of intraoperative electrostimulation parameters in
imaging of bipolar cortical stimulation. Neurosurg 78, conscious individuals: language cortex. J Neurosurg 15, 1–12.
785–793. Roux FE, Miskin K, Durand JB, Sacko O, Réhault E, Tanova R
Iwamura Y, Tanaka M, & Hikosaka O (1993). Rostrocaudal & Démonet JF (2015). Electrostimulation mapping of
gradients in the neuronal receptive field complexity in the comprehension of auditory and visual words. Cortex 71,
finger region of the alert monkey’s postcentral gyrus. Exp 398–408.
Brain Res 92, 360–368. Sanchez-Panchuelo RM, Besle J, Beckett A, Bowtell R,
Jones EG, Coulter JD & Hendry SH (1978). Intracortical Schluppeck D & Francis S (2012). Within-digit functional
connectivity of architectonic fields in the somatic sensory, parcellation of Brodmann areas of the human primary
motor and parietal cortex of monkeys. J Comp Neurol 181, somatosensory cortex using functional magnetic resonance
291–347. imaging at 7 tesla. J Neurosci 32, 15815–15822.
Kaas JH, Nelson RJ, Sur M, Lin CS & Merzenich MM (1979). Shoham D & Grinvald A (2001). The cortical representation of
Multiple representations of the body within the primary the hand in macaque and human area S-I: high resolution
somatosensory cortex of primates. Science 204, 521–3. optical imaging. J Neurosci 21, 6820–6835.
Kandel ER, Schwartz JH, Jessell TM, Siegelbaum SA & Urasaki E, Uematsu S, Gordon B & Lesser RP (1994). Cortical
Hudspeth AJ (2013). Principles of Neural Science, Fifth edn. tongue area studied by chronically implanted subdural
McGraw-Hill, New York. electrodes – with special reference to parietal motor and
Kolasinski J, Makin TR, Jbabdi S, Clare S, Stagg CJ & Johansen- frontal sensory responses. Brain 117, 117–132.
Berg H (2016). Investigating the stability of fine-grain digit Van Essen DC (2005). A population-average, landmark- and
somatotopy in individual human participants. J Neurosci 36, surface-based (PALS) atlas of human cerebral cortex.
1113–1127. Neuroimage 28, 635–662.
Krause T, Kurth R, Ruben J, Schwiemann J, Villringer K, Van Essen DC, Dickson J, Harwell J, Hanlon D, Anderson CH
Deuchert M, Moosmann M, Brandt S, Wolf K, Curio G & & Drury HA (2001). An integrated software system for
Villringer A (2001). Representational overlap of adjacent surface-based analyses of cerebral cortex. J Am Med Inform
fingers in multiple areas of human primary somatosensory Assoc 8, 443–459.
cortex depends on electrical stimulus intensity: an fMRI Vogt C & Vogt O (1919). Allgemeinere Ergebnisse unserer
study. Brain Res 899, 36–46. Hirnforschung. J Psychol Neurol 25, 279–462.
Martuzzi R, van der Zwaag W, Farthouat J, Gruetter R & Weinstein S (1968). Intensive and extensive aspects of tactile
Blanke O (2014). Human finger somatotopy in areas 3b, 1, sensitivity as a function of body part, sex and laterality. In
and 2: a 7T fMRI study using a natural stimulus. Hum Brain The Skin Senses, ed. Kenshalo DR, pp. 195–222. Charles
Mapp 35, 213–226. Thomas, Springfield, IL, USA.
Merzenich MM, Kaas JH, Sur M & Lin CS (2014). Double White LE, Andrews TJ, Hulette C, Richards A, Groelle M,
representation of the body surface within cytoarchitectonic Paydarfar J & Purves D (1997). Structure of the human
areas 3b and 1 in “SI” in the owl monkey (Aotus trivirgatus). sensorimotor system. I: Morphology and cytoarchitecture
J Comp Neurol 181, 41–73. of the central sulcus. Cereb Cortex 7, 18–30.


C 2017 The Authors. The Journal of Physiology 
C 2017 The Physiological Society
16 F.-E. Roux and others J Physiol 000.00

Additional information Author contributions

Experiments were performed at The University of Toulouse III.
Competing interests
F.E.R. was responsible for the study design. F.E.R. and I.D. carried
The authors declare that they have no competing interests. out the experiments. F.E.R., I.D. and J.B.D. analysed and inter-
preted the data. F.E.R. and J.B.D. drafted the paper. All authors
revised the article critically for important intellectual content.
Funding All authors have approved the final version of the manuscript and
agree to be accountable for all aspects of the work. All persons
This work was supported by the CNRS (Centre National pour la designated as authors qualify for authorship, and all those who
Recherche Scientifique). qualify for authorship are listed.


C 2017 The Authors. The Journal of Physiology 
C 2017 The Physiological Society