Near-infrared spectroscopy (NIRS) in cognitive neuroscience of the
Joaquı ´n Fuster,TMichael Guiou, Allen Ardestani, Andrew Cannestra, Sameer Sheth,
Yong-Di Zhou, Arthur Toga, and Mark BodnerT
UCLA Neuropsychiatric Institute, Room 38-159, 760 Westwood Plaza, Los Angeles, CA 90095-1759, USA
Received 7 September 2004; revised 22 December 2004; accepted 20 January 2005
Available online 19 March 2005
We describe the use of near-infrared spectroscopy (NIRS) as a suitable
means of assessing hemodynamic changes in the cerebral cortex of
awake and behaving monkeys. NIRS can be applied to animals
performing cognitive tasks in conjunction with electrophysiological
methods, thus offering the possibility of investigating cortical neuro-
vascular coupling in cognition. Because it imposes fewer constraints on
behavior than fMRI, NIRS appears more practical than fMRI for
certain studies of cognitive neuroscience on the primate cortex. In the
present study, NIRS and field potential signals were simultaneously
recorded from the association cortex (posterior parietal and prefrontal)
of monkeys performing two delay tasks, one spatial and the other non-
spatial. Working memory was accompanied by an increase in oxy-
genated hemoglobin mirrored by a decrease in deoxygenated hemoglo-
bin. Both the trends and the amplitudes of these changes differed by
task and by area. Field potential records revealed slow negative
potentials that preceded the task trials and persisted during their
memory period. The negativity during that period was greater in
prefrontal than in parietal cortex. Between tasks, the potential differ-
ences were less pronounced than the hemodynamic differences. The
present feasibility study lays the groundwork for future correlative
studies of cognitive function and neurovascular coupling in the primate.
D 2005 Elsevier Inc. All rights reserved.
Keywords: Near-infrared spectroscopy; Cognitive neuroscience; Working
memory; Slow potentials
Near infrared spectroscopy (NIRS) allows the monitoring of
changes in oxygenation of hemoglobin (Hb) in the brain (Jo ¨bsis,
1977). NIRS has been applied to the study of normal and abnormal
brain function in several species, including rat (Wolf et al., 1997),
pig (Brun et al., 1997), cat (Cairns et al., 1986), and human
(Cannestra et al., 2003; Obrig and Villringer, 2003). While not a
direct measure of neuronal activity, NIRS, like other oxygenation-
based techniques such as optical imaging in the visible spectrum
and fMRI, relies on the tight coupling between changes in cerebral
electrical activity and local blood oxygenation. Whereas this
relationship, termed neurovascular coupling, has been shown to
be roughly linear by a variety of methods (Arthurs and Boniface,
2003; Logothetis et al., 2001; Mathiesen et al., 2000; Ngai et al.,
1999), recent studies question that linearity (Devor et al., 2003;
Mechelli et al., 2001; Obrig et al., 2002; Sheth et al., 2004).
Furthermore, recent data from the rat suggest that anesthetics
significantly alter the coupling relationship (Berwick et al., 2002;
Sicard et al., 2003), bringing into question results from anesthe-
tized preparations. The present study explores the potential of
NIRS, combined with electrical recording, for characterizing
oxygenation changes in higher cortical regions of monkeys
performing cognitive tasks.
During the performance of a working memory task, neuronal
assemblies have been shown by microelectrode studies to be active
in large regions of both frontal and posterior (post-central) cortex
(Chafee and Goldman-Rakic, 1998; Quintana and Fuster, 1999). In
both frontal and posterior regions, neuronal ensembles are attuned
to various task-related events, including the retention of a sensory
item for prospective action, as well as task-specific items such as
the chromatic or spatial attributes of a visual memorandum (Fuster
et al., 1982; Gnadt and Andersen, 1988). Those neuronal
populations appear directly involved in the performance of
working memory tasks inasmuch as the disruption of their function
leads to deficits in such performance. For example, a delayed
matching-to-sample (DMS) task with colors can be impaired by the
reversible inactivation–by local cooling–of either the lateral
prefrontal cortex (Quintana and Fuster, 1993) or the inferotemporal
cortex (Fuster et al., 1981). Functional imaging reveals that the
performance of working memory tasks induces the activation of
prefrontal cortical areas and, in addition, primary or secondary
sensory areas of posterior cortex (Courtney et al., 1996; D’Esposito
et al., 1995; Owen et al., 1996; Ranganath et al., 2004; Swartz et
al., 1995). A recent NIRS study (Tsujimoto et al., 2004) shows
1053-8119/$ - see front matter D 2005 Elsevier Inc. All rights reserved.
T Corresponding authors. Fax: +1 310 825 6792.
E-mail addresses: firstname.lastname@example.org (J. Fuster)8 email@example.com
Available online on ScienceDirect (www.sciencedirect.com).
NeuroImage 26 (2005) 215–220
hemodynamic activation in the prefrontal cortex of humans during
The main objective of the present study was to determine the
feasibility of using NIRS and electrophysiological methods,
together, in monkeys performing cognitive tasks. If that feasibility
was established, it would open up the potential of NIRS for the
study of neurovascular coupling in animal cognition, which is
currently difficult with other imaging methods. The present
application of NIRS focuses on oxygenation changes during
working memory performance. Two areas of the cortex, prefrontal
and posterior parietal, were examined by NIRS and contrasted with
each other in two visual memory tasks, one in which the
memorandum was spatial location and the other color. A special
focus of the study was to uncover relationships between NIRS
signals and cortical field potentials during the short-term retention
of sensory stimuli.
One adult, male rhesus monkey (Macaca mulatta) weighing
approximately 10 kg was used for this study. The animal was
housed and handled in accordance with NIH guidelines for animal
use and protocols approved by the UCLA Chancellor’s Animal
Research Committee. Prior to behavioral training and NIRS
imaging, the animal was placed under anesthesia in a special
stereotaxic instrument and subjected to a cranial MRI scan to
determine the precise coordinates of the cortical regions of interest
for NIRS imaging.
Working memory tasks
The monkey, sitting in a primate chair and facing a panel with
three translucent disks, was trained to perform two working
memory tasks with visual cues: delayed matching-to-sample
(DMS) and delayed response (DR). A schematic illustration of
the tasks is shown in Fig. 1. The first, DMS, is a non-spatial delay
task that uses colors as cues; the second, DR, is a spatial delay task
that uses stimulus positions as cues. In both tasks, the animal holds
its hand on a hand-rest at all times except for making a choice
(electronic sensors signal the chosen disk). At the beginning of
each DMS trial, the monkey is presented for 2 s with a central
colored disk, red or green (the sample). After a delay period of 20 s,
the two lateral disks become simultaneously lit, one with red and
the other with green light. The animal, for reward, is then required
to choose the disk with the sample color. In DR, the cue that begins
the trial is a white-lit disk, on the right or on the left. After the 20-s
delay, the two lateral disks are simultaneously lit with white light,
and the animal then has to choose the disk on the side of the cue. In
both tasks, the color or position of the cue is changed at random
from trial to trial, and so is the position of the correct disk at the
time of choice. Therefore, both tasks are tests of working memory,
one spatial (DR) and the other non-spatial (DMS). In the present
study, trials were presented approximately every 80 s.
After training in the two tasks, the animal was subjected to
surgical preparation under general anesthesia. Two hollow stain-
less-steel cylinders (19 mm internal diameter) were implanted over
posterior parietal and lateral prefrontal cortex through trephine
holes above the intact dura. Fig. 2 shows the position of the two
cylinders with respect to cortical landmarks. Each cylinder was
anchored with acrylic cement and the ensemble reinforced with
small stainless steel screws inserted in the skull. Four threaded
metal sockets were embedded in the cement for head fixation with
external brackets during recording procedures. When not in use for
recording, the two cylindrical chambers were filled with solid
stainless steel cylinders and capped.
For a recording session, the two cylinders were opened and
filled with a snuggly-fitting cylindrical plastic insert containing
in its core the NIRS probes. These consisted of the ends of two
fiber-optic cables (boptodesQ), one the emitter and the other the
detector—1 cm apart from each other. With the insert in place,
the ends of the optodes were in direct contact with the dura.
Incident light was provided, via the emitter optode, by a stable
voltage-regulated halogen source, filtered between 650 and 1000
nm (Oriel Instruments). The detector optode collected light after
it had been scattered through the underlying cortex and guided it
to a spectrophotometer (SpectraPro 3001, Roper Scientific). The
reflected light, not absorbed by tissue, was dispersed over a
wavelength range of 730–869 nm and scanned by a CCD camera
(Spec-10, 400 R, Roper Scientific) equipped with a high-speed
shutter. Spectra were acquired at the rate of 3 Hz. Continuous
records were obtained beginning 10 s before onset of the cue
for each trial and ending 25 s after the animal’s behavioral
response that terminated the trial. The input from a movement
detector and transducer was used to screen out trials with exces-
sive movement. The animal’s behavior was monitored during the
recording sessions by means of another camera and a TV
Concomitant with the recording of NIRS signals, silver-ball
epidural electrodes were used to record field potentials from the
regions of interest. Field potentials were recorded monopolarly
(reference earlobe) and amplified by means of Grass amplifiers
(Mod. 15 RX) with band-pass filtering set at 0.01–30 Hz.
Fig. 1. Diagram of the two delay tasks. Delayed matching-to-sample (DMS)
required the memorization of color, and delayed response (DR) the
memorization of position, through a 20-s delay. Abbreviation: c, correct
J. Fuster et al. / NeuroImage 26 (2005) 215–220
The concentration changes of oxy-hemoglobin [HbO2] and
deoxy-hemoglobin [Hbr] were calculated using a modified Beer–
Lambert law (described by Uludag et al., 2002):
I t ð Þk¼ Ek
where Iois incident light intensity, I(t) detected light intensity at
time t, k wavelength, E extinction coefficient for the chromo-
phores, and L pathlength. Loss of light due to scatter changed
minimally during recording and could therefore be disregarded
when assessing changes in attenuation (Obrig and Villringer,
2003). Wavelength dependency (indicated by k) for both the
extinction coefficients and the optical pathlengths were determined
by Monte Carlo simulation and the diffusion approximation (Kohl
et al., 2000; Wray et al., 1988). Briefly, the Beer–Lambert law
allows the measurement of the attenuation of light intensity by
absorption of chromophores (oxy- and deoxy-hemoglobin) in the
volume of illuminated tissue.
Both NIRS and field potential records were selected for correct
performance by the animal and freedom from artifacts. The records
were averaged across trials of a given task for each experimental
session. Then the averages were compared for differences as a
function of task and cortical site. Activation from inter-trial
baseline levels of [HbO2] and [Hbr] was determined using
Student’s t test (two-tailed) to compare post-cue time points to
baseline values. Deviations from baseline with a P value less than
0.05 were considered statistically significant. The analysis allowed
the assessment–within region–of relative hemodynamic changes in
the course of each task, as well as the examination of temporal
correlations between NIRS signals and surface field potentials
during the monkey’s performance of the tasks.
½? t ð Þ þ Ek
½? t ð Þ
As measured by NIRS, the hemodynamic response of neural
tissue to a stimulus consists of an increase in [HbO2] and a
decrease in [Hbr]. The magnitude of the increase in [HbO2] is
typically two to three times that of the decrease in [Hbr]. Thus, the
net result is an increase in total oxygenated hemoglobin (Obrig and
Villringer, 2003). Similar trends were observed here at the
presentation of the cue or memorandum in either of the two tasks,
DR and DMS. Those trends persisted throughout the task trial,
although some differences were observed in the relative magnitude
of the changes depending on the task and the cortical location.
Figs. 3 and 4 depict NIRS response averages for each of the two
cortical sites and memory tasks.
In both cortical regions and tasks, [HbO2] was elevated above
inter-trial baseline in the course of the delay as well as in the
subsequent choice period. The increase began at or before the cue
and reached a relatively high level around the middle or latter part
of the 20-s delay. After a subsequent descent that persisted into the
beginning of the choice period, [HbO2] resumed its increase and
reached a second elevation in the choice period. In the parietal
cortex, the delay maximum was more pronounced–and occurred
later–in the DMS task than in the DR task (0.9% at 19 s versus
0.5% at 10 s, respectively). Conversely, in the prefrontal cortex, the
delay maximum was more pronounced in the DR task than in DMS
task (0.45% at 15 s versus 0.25% at 19 s, respectively). Whereas in
the parietal cortex oxygenation changes during the choice period
Fig. 3. NIRS and field potential records from the posterior parietal cortex
during delay tasks: delayed response (DR) and delayed matching-to-sample
(DMS). For the purpose of averaging across trials and sessions, records are
time-locked with cue and choice stimuli presentation. (Top) Average
changes in [HbO2] and [Hbr]. Error bars are measures of inter-session
variance. (Bottom) Averaged field potentials.
Fig. 2. Lateral view of the brain of the monkey with approximate
demarcation–in red–of the two cortical regions, posterior parietal (area
7a) and prefrontal (area 9), from which NIRS signal and field potentials
were recorded. Abbreviations: as, arcuate sulcus; cs, central sulcus; ips,
intraparietal sulcus; ls, lateral sulcus; ps, principal sulcus; sts, superior
J. Fuster et al. / NeuroImage 26 (2005) 215–220
were generally larger than during the delay, the reverse seemed to
occur in the prefrontal cortex. Changes in [HbO2] were accom-
panied by opposite and slightly protracted, lesser magnitude,
changes in [Hbr].
In addition to the large, persistent changes noted in the delay
and choice periods, transient and less conspicuous changes were
observed before and during the cue or memorandum that started
each trial. Because in both tasks the trials were administered at
somewhat regular intervals, the animal could predict approximately
the time of the cue. Probably because of it, in both cortical regions
a slight increase in oxygenation preceded the cue by some 5 s
(Figs. 3 and 4).
In both cortical regions, very slow field potentials were
recorded during inter-trial periods. A negativity of about 30–40
AV was commonly detected in both locations before trials (Figs. 3
and 4). The cue induced a triphasic positive–negative–positive
evoked potential (the first positive component small and com-
monly undetectable), followed by a resumption of negativity that
persisted for the duration of the delay. For most of the delay,
sustained negativity was more prominent in prefrontal than in
parietal cortex. Task-dependent differences were less apparent than
area-dependent differences. In both tasks and areas, the events in
the choice period, including the choice stimuli and the motor
response of the animal, were accompanied by a large negative–
Our study shows that NIRS is a suitable method to measure
oxygenation changes in the cortex of the monkey performing
cognitive tasks. The method is only moderately invasive, in that it
can detect those changes through the dura and without the
necessity to remove it. However, whereas NIRS has excellent
temporal resolution, its spatial resolution is inferior to that of fMRI.
In terms of resolution, therefore, NIRS offers no advantage over
fMRI, which has also been successfully used in monkeys
performing cognitive tasks (Koyama et al., 2004; Nakahara et
al., 2002). NIRS, however, seems to be more compatible than
fMRI with both the normal behavior of the monkey in a
comfortable sitting position and the concomitant electrical record-
ing of local potentials from its cerebral cortex.
Compared to other optical imaging methods, NIRS has definite
limitations. Notably, it does not possess the spatial definition of
optical methods with light in the visible spectrum. Therefore, it is
not ideal for precise analysis of surface cortical hemodynamics. At
the same time, because of NIRS’ greater tissue penetrance, it
allows the detection of oxygenation changes in the depth of the
cortex, although it does not have the degree of penetrance that
fMRI has for subcortical imaging. Nonetheless, for applications
such as those of the current study in the monkey, NIRS appears to
be a more convenient method than fMRI.
The most salient finding of our study was a significant increase
in cortical hemodynamic activity during the retention of informa-
tion in active short-term memory. In both tasks and cortical areas,
we observed an increase in [HbO2] during the delay phase of the
task, accompanied by a nearly symmetrical decrease in [Hbr].
These changes are comparable to those observed by Tsujimoto et
al. (2004) using NIRS in the prefrontal cortex of the human. In the
monkey, the increase in oxygenation during working memory was
concomitantly accompanied by slow negative potentials, especially
prominent in prefrontal cortex. The [HbO2] increase during the
delay probably results from the recruitment of large populations of
cortical cells engaged in working memory. Cellular recruitment is
reflected by the negative components of the field potentials during
the delay. The activated cell populations at the source of those
potentials may participate in working memory as well as in the
preparation for behavioral response (Quintana and Fuster, 1999). In
any case, the neuronal interpretation of hemodynamic differences
between tasks and regions must await the investigation of cellular
discharge from the same regions in identical behavioral conditions.
The transient, small-scale changes we observed in certain task
periods may correspond to discrete changes observed in compa-
rable neurophysiological or imaging experiments. For example, the
[HbO2] increase preceding a trial probably reflected the antici-
patory discharge of prefrontal and parietal neurons before a
stimulus whose timing the animal could predict (Quintana and
Fuster, 1999; Watanabe, 1986).
By simultaneously recording BOLD fMRI and electrical signals
from the visual cortex of the anesthetized monkey, Logothetis et al.
Fig. 4. NIRS and field potential records from the prefrontal cortex during
delay tasks. Same conventional notations are used as in Fig. 3.
J. Fuster et al. / NeuroImage 26 (2005) 215–220
(2001) showed that slow changes in local field potentials–more so
than cell discharge–were temporally correlated with BOLD
responses. The present study shows comparable temporal relation-
ships between NIRS and surface field potentials in higher
association cortices during cognitive performance, while also
revealing region-differential trends of surface potential negativity
and increased oxygenated blood during working memory. It
remains to be seen whether similar relationships can be observed
between NIRS and cell discharge in higher cortices of the
The results of this methodological study hold promise for future
studies of the neurovascular correlates of cognitive function in
high-order neocortex. The NIRS method is eminently suited for the
study of cognitive function in the non-human primate. It is
relatively economical, safe, noiseless, and compatible with both
instrumental behavior and electrophysiological recording. The use
of NIRS in a multi-method approach may also be of clinical
interest. Several models of degenerative cognitive diseases such as
Alzheimer’s have been developed in the primate (Lane, 2000;
Voyko, 1998). Recent studies in humans show that changes in
hemodynamic activity precede the onset of behavioral symptoms
in those diseases (Burggren et al., 2002). Using a combined
vascular and electrophysiological approach, it should be possible to
better assess and predict disease progression. Similarly, this
approach could also be of valuable for assessing efficacy of
We thank Bradford Lubell and William Bergerson for their
technical assistance and Arno Villringer and Helmuth Obrig for
their advise in the assembly of our NIRS equipment. This work
was supported by NSF grant IBN-9905053 and NIMH grants MH-
51697 and MH-52083.
Arthurs, O.J., Boniface, S., 2003. What aspect of the fMRI BOLD signal
best reflects the underlying electrophysiology in human somatosensory
cortex? Clin. Neurophysiol. 114, 1203–1209.
Berwick, J., Martin, C., Martindale, J., Jones, M., Johnson, D., Zheng, Y.,
Redgrave, P., Mayhew, J., 2002. Hemodynamic response in the
unanesthetized rat: intrinsic optical imaging and spectroscopy of the
barrel cortex. J. Cereb. Blood Flow Metab. 22, 670–679.
Brun, N.C., Moen, A., Borch, K., Saugstad, O.D., Greisen, G., 1997. Near-
infrared monitoring of cerebral tissue oxygen saturation and blood
volume in newborn piglets. Am. J. Physiol. 273, H682–H686.
Burggren, A.C., Small, G.W., Sabb, F., Bookheimer, S.Y., 2002. Specificity
of brain activation patterns in people at genetic risk for Alzheimer
disease. Am. J. Geriatr. Psychiatry 10, 44–51.
Cairns, C.B., Fillipo, D., Palladino, G.W., Proctor, H., 1986. Direct
noninvasive assessment of brain metabolism during increased intra-
cranial pressure: potential therapeutic vistas. J. Trauma 26, 863–868.
Cannestra, A.F., Wartenburger, I., Obrig, H., Villringer, A., Toga, A.W.,
2003. Functional assessment of Broca’s area using near infrared
spectroscopy in humans. NeuroReport 14, 1961–1965.
Chafee, M.V., Goldman-Rakic, P.S., 1998. Matching patterns of
activity in primate prefrontal area 8a and parietal area 7ip neu-
rons during a spatial working memory task. J. Neurophysiol. 79,
Courtney, S.M., Ungerleider, L.G., Keil, K., Haxby, J.V., 1996. Object and
spatial visual working memory activate separate neural systems in
human cortex. Cereb. Cortex 6, 39–49.
D’Esposito, M., Detre, J.A., Alsop, D.C., Shin, R.K., Atlas, S., Grossman,
M., 1995. The neural basis of the central executive system of working
memory. Nature 378, 279–281.
Devor, A., Dunn, A.K., Andermann, M.L., Ulbert, I., Boas, D.A., Dale,
A.M., 2003. Coupling of total hemoglobin concentration, oxygen-
ation, and neural activity in rat somatosensory cortex. Neuron 39,
Fuster, J.M., Bauer, R.H., Jervey, J.P., 1981. Effects of cooling infer-
otemporal cortex on performance of visual memory tasks. Exp. Neurol.
Fuster, J.M., Bauer, R.H., Jervey, J.P., 1982. Cellular discharge in the
dorsolateral prefrontal cortex of the monkey in cognitive tasks. Exp.
Neurol. 77, 679–694.
Gnadt, J.W., Andersen, R.A., 1988. Memory related motor planning
activity in parietal cortex of macaque. Exp. Brain Res. 70, 216–220.
Jfbsis, F.F., 1977. Noninvasive, infrared monitoring of cerebral and
myocardial oxygen sufficiency and circulatory parameters. Science
Kohl, M., Lindauer, U., Royl, G., Kuhl, M., Gold, L., Villringer, A.,
Dirnagl, U., 2000. Physical model for the spectroscopic analysis of
cortical intrinsic optical signals. Phys. Med. Biol. 45, 3749–3764.
Koyama, M., Hasegawa, I., Osada, T., Adachi, Y., Nakahara, K., Miyashita,
Y., 2004. Functional magnetic resonance imaging of macaque monkeys
performing visually guided saccade tasks: comparison of cortical eye
fields with humans. Neuron 41, 795–807.
Lane, M., 2000. Nonhuman primate models in biogerontology. Exp.
Gerontol. 35, 533–541.
Logothetis, N.K., Pauls, J., Augath, M., Trinath, T., Oeltermann, A., 2001.
Neurophysiological investigation of the basis of the fMRI signal.
Nature 412, 150–157.
Mathiesen, C., Caesar, K., Lauritzen, M., 2000. Temporal coupling between
neuronal activity and blood flow in rat cerebellar cortex as indicated by
field potential analysis. J. Physiol. 523, 235–246.
Mechelli, A., Price, C.J., Friston, K.J., 2001. Nonlinear coupling between
evoked rCBF and BOLD signals: a simulation study of hemodynamic
responses. NeuroImage 14, 862–872.
Nakahara, K., Hayashi, T., Konishi, S., Miyashita, Y., 2002. Functional
MRI of macaque monkeys performing a cognitive set-shifting task.
Science 22, 1532–1536.
Ngai, A.C., Jolley, M.A., D’Ambrosio, R., Meno, J.R., Winn, H., 1999.
Frequency-dependent changes in cerebral blood flow and evoked
potential during comatose somatosensory in the rat. Brain Res. 837,
Obrig, H., Villringer, A., 2003. Beyond the visible—imaging the human
brain with light. J. Cereb. Blood Flow Metab. 23, 1–18.
Obrig, H., Israel, H., Kohl-Bareis, M., Uludag, K., Wenzel, R., Muller, B.,
Arnold, G., Villringer, A., 2002. Habituation of the visually evoked
potential and its vascular response: implications for neurovascular
coupling in the healthy adult. NeuroImage 17, 1–18.
Owen, A.M., Morris, G., Sahakian, B.J., Polkey, C.E., Robbins, T.W.,
1996. Double dissociations of memory and executive functions in
working memory tasks following frontal lobe excisions, temporal
lobe excisions or amygdalo-hippocampectomy in man. Brain 119,
Quintana, J., Fuster, J.M., 1993. Spatial and temporal factors in the role of
prefrontal and parietal cortex in visuomotor integration. Cereb. Cortex
Quintana, J., Fuster, J.M., 1999. From perception to action: temporal
integrative functions of prefrontal and parietal neurons. Cereb. Cortex 9,
Ranganath, C., Cohen, M., Dam, C., D’Esposito, M., 2004. Inferior
temporal, prefrontal, and hippocampal contributions to visual working
memory maintenance and associative memory retrieval. J. Neurosci. 24,
Sheth, S., Nemoto, M., Guiou, M., Walker, M., Pouratian, N., Toga, A.W.,
J. Fuster et al. / NeuroImage 26 (2005) 215–220
2004. Linear and nonlinear relationships between neuronal ac- Download full-text
tivity, oxygen metabolism, and hemodynamic responses. Neuron 42,
Sicard, K., Shen, Q., Brevard, M.E., Sullivan, R., Ferris, C.F., King, J.A.,
Duong, T.Q., 2003. Regional cerebral blood flow and BOLD responses
in conscious and anesthetized rats under basal and hypercapnic
conditions: implications for functional MRI studies. J. Cereb. Blood
Flow Metab. 23, 472–481.
Swartz, B.E., Halgren, E., Fuster, J.M., Simpkins, F., Gee, M., Mandelkern,
M., 1995. Cortical metabolic activation in humans during a visual
memory task. Cereb. Cortex 3, 205–214.
Tsujimoto, S., Yamamoto, T., Kawaguchi, H., Koizumi, H., Sawaguchi, T.,
2004. Prefrontal cortical activation associated with working memory in
adults and preschool children: an event-related optical topography
study. Cereb. Cortex 14, 703–712.
Uludag, K., Kohl, M., Steinbrink, J., Obrig, H., Villringer, A., 2002. Cross
talk in the Lambert–Beer calculation for near-infrared wavelengths
estimated by Monte Carlo simulations. J. Biomed. Opt. 7, 51–59.
Voyko, M., 1998. Nonhuman primates as models for aging and Alzheimer’s
disease. Lab. Anim. Sci. 48, 611–617.
Watanabe, M., 1986. Prefrontal unit activity during delayed conditional go/
no-go discrimination in the monkey: I. Relation to the stimulus. Brain
Res. 382, 1–14.
Wolf, T., Lindauer, U., Reuter, U., Back, T., Villringer, A., Einhaupl, K.,
Dirnagl, U., 1997. Noninvasive near infrared spectroscopy monitoring
of regional cerebral blood oxygenation changes during peri-infract
depolarizations in focal cerebral ischemia in the rat. J. Cereb. Blood
Flow Metab. 17, 950–954.
Wray, S., Cope, M., Delpy, D.T., Wyatt, J.S., Reynolds, E.O.R., 1988.
Characterization of the near-infrared absorption spectra of cytochrome
aa3 and haemoglobin for the non-invasive monitoring of cerebral
oxygenation. Biochim. Biophys. Acta 933, 184–192.
J. Fuster et al. / NeuroImage 26 (2005) 215–220