Neuroscience Letters 451 (2009) 129–133
Contents lists available at ScienceDirect
Neuroscience Letters
journal homepage: www.elsevier.com/locate/neulet
Activation in ventro-lateral prefrontal cortex during the act of tasting:
An fNIRS study
Masako Okamoto ∗ , Haruka Dan, Lester Clowney, Yui Yamaguchi, Ippeita Dan ∗
National Food Research Institute, 2-1-12 Kannondai, Tsukuba 305-8642, Japan
a r t i c l e
i n f o
Article history:
Received 10 September 2008
Received in revised form 5 December 2008
Accepted 10 December 2008
Keywords:
Optical topography
Taste
Oral-sensation
Feeding behavior
Obesity
Eating disorders
a b s t r a c t
The act of tasting is the product of inseparable integrative behavior consisting of multi-sensory processing
and orolingual motor coordination. Often tasting-induced brain activity is looked at in a reductionist
manner as a set of isolated components. However, brain activity as a whole during tasting may not simply
be the sum of isolated brain responses; therefore, attempting to look at the cortical activation in a more
holistic manner is important. Using functional near-infrared spectroscopy (fNIRS), we assessed cortical
responses during tasting, contrasting observed neuronal activation of the lateral prefrontal cortex (LPFC),
of 19 healthy participants before and during tasting of 8 ml of sweet-based solutions. To examine the
activated brain structure, we estimated the anatomical regions of the measured location in standard
brain space. We also included simple tongue tapping movement (TT) and word fluency (WF) tasks as
comparative functional markers. Significant activity was found in channels (CHs) estimated to be in the
bilateral oral motor areas during the TT task, and those in the LPFC, primarily in the left hemisphere,
during the WF task. During the tasting task, significant activation was observed in CHs estimated to lie in
the ventral part of pre- and post-central gyri as well as in the ventro-LPFC (VLPFC). The activated regions
partly overlapped with those detected during TT or WF tasks, but extended more anteriorly and ventrally.
Our study suggests that, in addition to tongue motor areas, the VLPFC is involved in the act of tasting.
© 2008 Elsevier Ireland Ltd. All rights reserved.
Tasting food involves not only taste processing, but also complex
multimodal processing including oral-somatosensation, olfaction,
oral motor functions, and some cognitive functions [18]. The central
processing related to these functions has been studied mainly by
assessing each effect separately. For example, taste-related brain
functions have been assessed by comparing the cortical activation
elicited by taste and tasteless solutions to exclude other sensory and
motor effects [10]. Thus, cortical regions involved in each element
related to tasting have been successfully elucidated [20]. However,
brain responses during tasting as a whole may not be the same as
their componential sum. Therefore, a more holistic approach may
be needed to understand the neural basis for feeding behaviors.
In this report, we explore lateral prefrontal (LPFC) activity during
tasting using functional near-infrared spectroscopy (fNIRS). To now,
the major human brain mapping techniques used for taste related
functions are positron emission tomography (PET), functional magnetic resonance imaging (fMRI), and magnetoencephalography
(MEG). These neuroimaging techniques often require participants
to be fixed in a supine position with strict movement restrictions,
and often necessitate specialized stimuli presentation methods
∗ Corresponding authors. Tel.: +81 29 838 7357; fax: +81 29 838 7319.
E-mail addresses: masakoo@affrc.go.jp (M. Okamoto), dan@affrc.go.jp (I. Dan).
URL: http://brain.job.affrc.go.jp/ (I. Dan).
0304-3940/$ – see front matter © 2008 Elsevier Ireland Ltd. All rights reserved.
doi:10.1016/j.neulet.2008.12.016
(e.g., stimuli presentation via a small hole in a tube to a specific part
of the tongue [12]). They are well designed to optimize the use of
each neuroimaging technique, but are very different from the conditions under which we taste in our everyday lives. In contrast, fNIRS,
an optical method that non-invasively measures cortical hemodynamic responses [11], is relatively forgiving of body movement, and
is less restrictive (Fig. 1A). In fNIRS measurements, participants simply wear a set of probes on their heads, and can, therefore, taste
samples in an upright position [13]. In this respect, fNIRS has better
potential for examining brain responses in a holistic manner while
tasting under more natural conditions than other neuroimaging
techniques.
fNIRS can measure the responses of lateral cortical surfaces, but
not gustatory areas such as the insular and orbitofrontal cortices,
which are located deep inside the brain where the near-infrared
light cannot reach. Among the regions where fNIRS can be used,
the LPFC is of interest, as a recent study has suggested that taste
and other food-related activities occur in this area [9]. Using fMRI,
Kringelbach et al. reported activity in the dorso-LPFC (DLPFC)
related to taste, and suggested that taste constantly evokes cognitive processes mediated by this region [9]. This finding evoked
research interest in fields associated with feeding behaviors, such
as obesity and eating disorders [26]. In our fNIRS experiments,
however, we consistently observe activities in more ventral areas,
in addition to the DLPFC, during tasting in contrast with resting
130
M. Okamoto et al. / Neuroscience Letters 451 (2009) 129–133
Fig. 1. Experiment setting. (A) One of the authors receiving a taste solution is
monitored using multi-CH fNIRS for demonstration. She is connected to the fNIRS
equipment behind her with optical fibers whose ends, consisting of illuminatordetector pairs, are attached to the surface of her head. (B) Schematic depiction of CH
arrangements for the left hemisphere. A CH (gray circles) is set between each pair
of detectors (gray squares) and illuminators (white squares) resulting in 17 CHs per
hemisphere. Orientations of the illuminator-detector pairs and the 10–10 marker
positions used as cranial landmarks are indicated.
conditions. If we widen our scope from pure taste processing to
the act of tasting as a whole, the involvement of larger LPFC areas
may be observed, yielding a more holistic picture of tasting.
LPFC areas, which include the motor language areas, are adjacent
to oral sensori-motor areas. As fNIRS does not provide anatomical information of the measured brain, we examined the locations
of tasting-related activities in relation to these well-known functional sub-regions by measuring brain activities of a simple tongue
tapping movement (TT) task as well as a word fluency (WF) task.
In addition, we estimated anatomical regions using a database by
probabilistically registering the measured location to the standard
brain space as previously described [19]. In this way, we sought to
characterize the activation location related to tasting measured by
fNIRS.
Nineteen participants (13 females; 6 males; average age = 32.1
years; S.D. = 6.9; range: 23–44 years) participated in this study. We
excluded four participants’ tasting data because of event-related
noise likely due to temporal muscle movements. One participant
could not attend the WF task session. Written informed consent
was obtained after a complete explanation of the study. The study
was approved by the National Food Research Institute’s institutional
ethics committee.
Participants performed three tasks in separate sessions: a tasting task, a word fluency (WF) task, and a tongue tapping (TT) task.
All the tasks were performed using a block design. In all the sessions, participants sat in a quiet room with their eyes closed during
the trials. An experimenter monitored the participants’ behavior
throughout, and the trials with event-related noise were removed
from analysis. The order of the sessions was randomized across
participants.
The WF and TT sessions shared the same time schedule, and
consisted of 10 trials each. For each trial, task blocks were 20 s, and
spaced 20–24 s apart. In WF sessions, the experimenter announced
a word from one of ten categories (e.g., fruit, countries, etc.) at the
onset, and participants were instructed to silently (internal speech)
think of as many nouns as possible in the given category during the
task period, stopping when instructed. In TT sessions, the participants touched their tongue to the back of each of their upper teeth
sequentially from the right back tooth to the left back tooth and
back again, at a rate of approximately 1.2 Hz during the task period.
The start and the end of the task period were indicated orally using
the words “start” and “stop.”
Tasting sessions consisted of 6 trials, each consisting of 40 s of
rest, 20 s of tasting, and 10 s of rinsing, with about 60 s inter-trial
intervals. We used fewer trials and longer rest/interval blocks than
the other tasks to avoid sensory fatigue, and to provide enough
time for fNIRS signals to stabilize before the baseline period after
the straw insertion or mouth rinsing [13]. At the beginning of the
rest period, an experimenter inserted the end of a straw attached to
a syringe into the participant’s mouth and manually injected 8 ml
of a tasting sample at the onset of the tasting period. Each participant held the sample in his/her mouth during the entire tasting
block and then spit it out to avoid satiety. We used a total of 8
sugar-based taste stimuli with slightly different concentrations of
sour, salty, and umami/savory tastes (Supplemental Table 1). The
stimuli were made so as to avoid overall bias toward specific characteristics (e.g., not all were intense, unpleasant, etc.) to minimize
and counter-balance the effect of emotional reactions. The selection of samples was counter-balanced evenly across participants,
the order of presentation was randomized between trials, and the
samples were served at room temperature. After the fNIRS session,
we asked each participant to taste the samples again, evaluating the
familiarity, pleasantness, and intensity of each. We used a scale of
1–5 (5 = very familiar/pleasant/strong, 3 = neutral, 1 = very unfamiliar/unpleasant/weak). The post-experiment ratings revealed that,
on average, taste samples were perceived as neutral regarding
pleasantness (mean rating score 3.0; S.D. = 0.9), intensity (mean
rating score 3.4; S.D. = 1.0), and familiarity (mean rating score 3.1;
S.D. = 1.0).
We used the fNIRS topography system OMM-2000 Optical
Multi-channel Monitor (Shimadzu, Kyoto, Japan), which uses nearinfrared light with wavelengths of 780, 805, and 830 nm. We set
6 illuminator-detector pairs (distance between illuminator and
detector = 3 cm) in a 3 × 4 lattice pattern to form 17 channels (CHs),
and placed one holder over the frontal region of each hemisphere,
using F7 (F8) and C5 (C6) of the international 10–10 system as reference points for holder placement (Figs. 1B and 2A). We analyzed
optical data based on the modified Beer–Lambert Law, and calculated signals reflecting the oxygenated hemoglobin (oxyHb) and
deoxygenated hemoglobin (deoxyHb) concentration changes, in an
arbitrary unit (millimolar–millimeter) as previously described [13].
Statistical analysis was performed using (task–rest) contrast for
each task using a random effects summary statistics approach as
previously described [13]. A boxcar function was employed at the
first level to generate participants’ contrasts for all CHs. At the
second level, the single sample t test (one-tail) was employed to
test increases of oxyHb and decreases of deoxyHb. The resulting pvalues from all CHs were thresholded using the false discovery rate
control (FDR) method (P < FDR 0.05) as previously described [13],
for multiple testing correction.
We used our probabilistic estimation method [19] to estimate
CH location to the Montreal neurological institute (MNI) standard
brain space. Briefly, fNIRS optode positions, together with several
scalp landmarks, were digitized using a 3D magnetic space digitizer
(FASTRAK – Polhemus, Colchester, VT). Referring to our structural
head MRI database, we performed a registration simulation and
estimated locations of given CHs and their estimation errors for our
participants (Fig. 2; Supplementary Fig. 1 and Supplementary Table
2). We anatomically labeled these locations using a Matlab function
(available at http://brain.job.affrc.go.jp), which reads anatomical
labeling information coded in a brain atlas constructed by Shattuck
et al. [17]. Based on this estimation, we selected CHs located on the
frontal lobe and the post-central gyri for the functional analysis.
We observed significant activations in different LPFC areas
across different tasks (Fig. 2). In the TT task, a significant increase
in oxyHb was observed in CHs located in bilateral pre- and postcentral gyri (P < FDR 0.05). In the WF task, CHs covering part of the
inferior frontal gyrus (IFG) and middle frontal gyrus (MFG), mainly
in the left hemisphere, exhibited a significant oxyHb increase
(P < FDR 0.05). In the tasting task, an increase of oxyHb was observed
in CHs covering part of the precentral gyrus, MFG, and IFG in both
hemispheres (P < FDR 0.05). Only the TT task induced a significant
M. Okamoto et al. / Neuroscience Letters 451 (2009) 129–133
131
Fig. 2. Results of fNIRS analyses (A–C). CH-wise results are shown on our template brain in MNI space. The CHs (filled circles) that exhibited significant oxyHb increase
(P < FDR 0.05) are colored according to the t-values, as shown in the color bar, while those below the threshold are indicated in gray. The CHs that exhibited significant
deoxyHb decrease (P < FDR 0.05) are indicated with open blue circles. Pink dotted lines show borders of anatomical regions. Marker positions of the international 10–10
system we used as reference points for holder placement are indicated in (A). The time courses of oxyHb (red lines) and deoxyHb (blue lines) from the CH that exhibited
the highest t-value based on oxyHb analysis in each experiment is shown on the right. Error bars indicate standard deviation across participants. Error bars for post-task
rinsing period (shaded area) that includes noise due to muscle movements are truncated. (D) Locations of activation foci referred to in the discussion are shown based on
MNI coordinate values reported in original papers. Activations that were neither located within 1 cm of the lateral brain surface nor reported with coordinate values are not
included. The activated CHs for the tasting condition are shown as filled pink circles.
decrease in deoxyHb, centering in the same CHs where we detected
a significant increase of oxyHb (P < FDR 0.05). For the WF and tasting tasks, in the CHs exhibiting significant oxyHb increase, deoxyHb
tended to decrease, but inter-subject variability was large and the
decrease was not significant in most of the CHs.
These results are in line with previous studies. Regarding location, the areas activated by the TT task correspond to those
identified as oral sensory-motor areas by meta-analysis of PET stud-
ies [4], and left lateralized activation in the LPFC region during the
WF task is in line with earlier fMRI studies using this task (earlier
studies are listed in [3]). Regarding the behavior of each Hb signal, simultaneous recordings of fNIRS and fMRI during motor tasks
have produced oxyHb and deoxyHb changes consistent with fMRI
signals in primary sensori-motor areas corresponding to the task
[24]. As with these studies, both oxyHb and deoxyHb changes were
observed in oral motor areas during our TT task. For language tasks,
132
M. Okamoto et al. / Neuroscience Letters 451 (2009) 129–133
fNIRS studies have detected increases of oxyHb in the left LPFC area
where fMRI signal increases have been detected, but a decrease
of deoxyHb was not always observed [2]. Similarly, oxyHb signal
changes were observed in the left LPFC during our WF task.
For the tasting task, the taste-rest contrast we examined
includes multiple effects: taste, oral somato-sensory, and oral
movement. As expected for this contrast, we detected oxyHb
increase in the same oral sensory-motor area as for the TT task [4]
(Fig. 2D blue asterisks) as well as in more ventral regions in both
hemispheres. Somatotopically, the pre- and post-central regions
ventral to the tongue region represent the region for the laryngopharyngeal organ [14]. Activations observed in the ventral parts
of pre- and post-central gyri are also consistent with activations
related to taste stimuli [12] (Fig. 2D, blue circle), and salivation [15].
The portion of activations observed in the DLPFC may be related
to the taste-related activation reported by Kringelbach et al. [9]
(Fig. 2D, blue square).
However, the areas where we found oxyHb increases extended
ventrally and anteriorly in the IFG and part of the MFG: areas not
often associated with oral sensori-motor, or taste processing, but
which are located over the frontal operculum, a structure known
to be involve in gustatory processing. It may be that activity in the
frontal operculum itself was observed; however, this is unlikely as it
is too deep to be measured using fNIRS [5]. As the bilateral activation
observed did not overlap completely with that occurring during the
WF task, internal speech evoked during tasting, if any, cannot be the
only factor for this activation.
Of the few taste-related studies reporting activation in the
ventro-LPFC (VLPFC) (including the IFG), one fMRI study reported
activation related to aversive taste [21] (Fig. 2D, blue triangle).
However, this does not seem to be the case here because subjective rating scores indicate that our taste stimuli were, on average,
neutral regarding their pleasantness. Other studies have included
brain response to electrically evoked taste [1], and to taste or flavor
stimuli in contrast to weak-tasting solutions [23] or water [6], but
activation foci reported in these studies appear to be deeper than
the lateral brain surface, according to the reported coordinates.
Taste-related VLPFC activation may not be detected using fMRI,
whose signal is considered to mainly reflect changes of deoxyHb.
However, this does not explain why PET and MEG studies have not
detected activation in this area. Also, regarding the VLPFC area,
changes of oxyHb, rather than deoxyHb, was reported to agree
more closely with fMRI results [27]. Therefore, the lack of significant deoxyHb changes does not sufficiently explain the difference
of our fNIRS results and earlier taste-related neuroimaging results.
Alternatively, the activation we observed may correspond to that
subtracted out in other taste studies using a different baseline to
define contrasts: a tasteless solution as opposed to a simple resting state. Interestingly, there is an fMRI study reporting activation
in the VLPFC (Fig. 2D, green circles) while participants tasted fruit
juice. The baseline in their study was taken to be a resting activity rather than activity elicited by a tasteless solution. This may
imply that this part of the brain is active when liquid is “tasted”
in the mouth regardless of whether the liquid has taste or not. The
VLPFC has been implicated in many other functions including integration of sensory information with the control of oral movements
[7], and multi-sensory integration for creating flavor representation
[22] (Fig. 2D, yellow circle). Although we cannot identify the specific
role of these areas in tasting, such multi-sensory and motor integration functions may be involved during tasting, and may explain
the activation we found.
In this study, there was only one CH where deoxyHb decreased
significantly during the tasting task. This is possibly due to higher
individual differences of behaviors of deoxyHb than that of oxyHb,
as shown by studies using both animals [8] and humans [16], especially in the prefrontal area (summarized by Cannestra et al., in
their introduction [2]). Considering our observations and these
reports, deoxyHb behavior may be rather complicated, leading to
the inconsistent trends we observed.As demonstrated, using fNIRS
we detected activation not only in the oral-motor area, but also in
the VLPFC when participants tasted 8 ml of aqueous taste solutions.
Although more experiments with wider experimental conditions
are needed to generalize our findings, our results indicate the
possible involvement of these areas in tasting when the scope is
broadened to include holistic taste processing. While it is important to examine the cortical process for each element separately,
in order to understand the neural basis for tasting, examining the
brain response to tasting as a whole is required. Such holistic tasting
effects might also be of interest in clinical areas related to feeding
behaviors such as obesity, eating disorders, or dysphasia: recently,
using fNIRS, Uehara found that VLPFC activity differed between
patients with eating disorders and healthy control participants during a word fluency task [25]. Whether such differences occur in
tasting tasks may be worth future exploration.The compactness
and flexibility of fNIRS allows for brain measurements under relatively natural tasting conditions, and for its use in clinical settings.
The current findings are a step towards understanding the role of
the LPFC in tasting, and fNIRS will further contribute to this issue,
including possible clinical applications.
Acknowledgements
We thank Ms Miho Imamura and Ms Saho Matsuda for preparation of the data, and Ms Melissa Noguchi for examination of the
manuscript. This study was supported in part by the Program for
Promotion of Basic Research Activities for Innovative Bioscience,
Health and Labor Sciences Research Grants, Research on Psychiatric
and Neurological Diseases and Mental Health, and grant-in-aids
from the Ministry of Education, Culture, Sport, Science, and Technology, Japan (18700625 awarded to MO and 18390404, 19650079
to ID).
Appendix A. Supplementary data
Supplementary data associated with this article can be found, in
the online version, at doi:10.1016/j.neulet.2008.12.016.
References
[1] M.A. Barry, J.C. Gatenby, J.D. Zeiger, J.C. Gore, Hemispheric dominance of cortical activity evoked by focal electrogustatory stimuli, Chem. Senses 26 (2001)
471–482.
[2] A.F. Cannestra, I. Wartenburger, H. Obrig, A. Villringer, A.W. Toga, Functional
assessment of Broca’s area using near infrared spectroscopy in humans, Neuroreport 14 (2003) 1961–1965.
[3] S.G. Costafreda, C.H. Fu, L. Lee, B. Everitt, M.J. Brammer, A.S. David, A systematic
review and quantitative appraisal of fMRI studies of verbal fluency: role of the
left inferior frontal gyrus, Hum. Brain Mapp. (2006).
[4] P.T. Fox, A. Huang, L.M. Parsons, J.H. Xiong, F. Zamarippa, L. Rainey, J.L. Lancaster,
Location-probability profiles for the mouth region of human primary motorsensory cortex: model and validation, Neuroimage 13 (2001) 196–209.
[5] Y. Fukui, Y. Ajichi, E. Okada, Monte Carlo prediction of near-infrared light propagation in realistic adult and neonatal head models, Appl. Opt. 42 (2003)
2881–2887.
[6] J.F. Gautier, K. Chen, A. Uecker, D. Bandy, J. Frost, A.D. Salbe, R.E. Pratley, M.
Lawson, E. Ravussin, E.M. Reiman, P.A. Tataranni, Regions of the human brain
affected during a liquid-meal taste perception in the fasting state: a positron
emission tomography study, Am. J. Clin. Nutr. 70 (1999) 806–810.
[7] J.D. Greenlee, H. Oya, H. Kawasaki, I.O. Volkov, O.P. Kaufman, C. Kovach, M.A.
Howard, J.F. Brugge, A functional connection between inferior frontal gyrus and
orofacial motor cortex in human, J. Neurophysiol. 92 (2004) 1153–1164.
[8] Y. Hoshi, N. Kobayashi, M. Tamura, Interpretation of near-infrared spectroscopy
signals: a study with a newly developed perfused rat brain model, J. Appl.
Physiol. 90 (2001) 1657–1662.
[9] M.L. Kringelbach, I.E. de Araujo, E.T. Rolls, Taste-related activity in the human
dorsolateral prefrontal cortex, Neuroimage 21 (2004) 781–788.
[10] J. O’Doherty, E.T. Rolls, S. Francis, R. Bowtell, F. McGlone, Representation of pleasant and aversive taste in the human brain, J. Neurophysiol. 85 (2001) 1315–1321.
M. Okamoto et al. / Neuroscience Letters 451 (2009) 129–133
[11] H. Obrig, A. Villringer, Beyond the visible—imaging the human brain with light,
J. Cereb. Blood Flow Metab. 23 (2003) 1–18.
[12] H. Ogawa, M. Wakita, K. Hasegawa, T. Kobayakawa, N. Sakai, T. Hirai, Y.
Yamashita, S. Saito, Functional MRI detection of activation in the primary gustatory cortices in humans, Chem. Senses 30 (2005) 583–592.
[13] M. Okamoto, M. Matsunami, H. Dan, T. Kohata, K. Kohyama, I. Dan, Prefrontal
activity during taste encoding: an fNIRS study, Neuroimage 31 (2006) 796–
806.
[14] W. Penfield, E. Boldrey, Somatic motor and sensory representation in the cerebral cortex as studied by electrical stimulation, Brain 15 (1938) 389–443.
[15] W. Penfield, H. Jasper, Epilepsy and the functional anatomy of the human brain,
Little, Brown and Co, Boston, 1954.
[16] H. Sato, Y. Fuchino, M. Kiguchi, T. Katura, A. Maki, T. Yoro, H. Koizumi, Intersubject variability of near-infrared spectroscopy signals during sensorimotor
cortex activation, J. Biomed. Opt. 10 (2005) 44001.
[17] D.W. Shattuck, M. Mirza, V. Adisetiyo, C. Hojatkashani, G. Salamon, K.L. Narr,
R.A. Poldrack, R.M. Bilder, A.W. Toga, Construction of a 3D probabilistic atlas of
human cortical structures, Neuroimage 39 (2007) 1064–1080.
[18] S.A. Simon, I.E. de Araujo, R. Gutierrez, M.A. Nicolelis, The neural mechanisms
of gustation: a distributed processing code, Nat. Rev. 7 (2006) 890–901.
[19] A.K. Singh, M. Okamoto, H. Dan, V. Jurcak, I. Dan, Spatial registration of multichannel multi-subject fNIRS data to MNI space without MRI, Neuroimage 27
(2005) 842–851.
133
[20] D.M. Small, Central gustatory processing in humans, Adv. Otorhinolaryngol. 63
(2006) 191–220.
[21] D.M. Small, M.D. Gregory, Y.E. Mak, D. Gitelman, M.M. Mesulam, T. Parrish, Dissociation of neural representation of intensity and affective valuation in human
gustation, Neuron 39 (2003) 701–711.
[22] D.M. Small, J. Prescott, Odor/taste integration and the perception of flavor, Exp.
Brain Res. (2005) 345–357.
[23] M. Smits, R.R. Peeters, P. van Hecke, S. Sunaert, A 3 T event-related functional
magnetic resonance imaging (fMRI) study of primary and secondary gustatory
cortex localization using natural tastants, Neuroradiology 49 (2007) 61–71.
[24] J. Steinbrink, A. Villringer, F. Kempf, D. Haux, S. Boden, H. Obrig, Illuminating
the BOLD signal: combined fMRI-fNIRS studies, Magn. Reson. Imag. 24 (2006)
495–505.
[25] T. Uehara, M. Fukuda, M. Suda, M. Ito, T. Suto, M. Kameyama, Y. Yamagishi,
M. Mikuni, Cerebral blood volume changes in patients with eating disorders
during word fluency: a preliminary study using multi-channel near infrared
spectroscopy, Eat Weight Disord. 12 (2007) 183–190.
[26] A. Wagner, H. Aizenstein, G.K. Frank, J. Figurski, J.C. May, K. Putnam, L. Fischer,
U.F. Bailer, S.E. Henry, C. McConaha, V. Vogel, W.H. Kaye, Neural correlates of
habituation to taste stimuli in healthy women, Psychiatry Res. 147 (2006) 57–67.
[27] T. Yamamoto, T. Kato, Paradoxical correlation between signal in functional magnetic resonance imaging and deoxygenated haemoglobin content in capillaries:
a new theoretical explanation, Phys. Med. Biol. 47 (2002) 1121–1141.