Korean J Physiol Pharmacol
Vol 17: 299－306, August, 2013
ABBREVIATIONS: PD, Parkinson’s disease; DBS, deep brain
stimulation; STN, subthalamic nucleus; GP, globus pallidus.
Received February 27, 2013, Revised May 16, 2013,
Accepted June 5, 2013
Corresponding to: Kyung Hwan Kim, Department of Biomedical
Engineering, College of Health Science, Yonsei University, 234
Maeji-ri, Heungup-myun, Wonju 220-710, Korea. (Tel) 82-33-760-
2932, (Fax) 82-33-763-1953, (E-mail) khkim0604@ yonsei.ac.kr
This is an Open Access article distributed under the terms of the
Creative Commons Attribution Non-Commercial License (http://
creativecommons.org/licenses/by-nc/3.0) which permits unrestricted non-commercial
use, distribution, and reproduction in any medium, provided the original work
is properly cited.
Neuronal Responses in the Globus Pallidus during Subthalamic
Nucleus Electrical Stimulation in Normal and Parkinson’s Disease
Sang Baek Ryu1, Eun Kyung Bae1, Jinhyung Kim2,3, Yong Sup Hwang2,3, Changkyun Im4, Jin Woo
Chang2,3, Hyung-Cheul Shin4, and Kyung Hwan Kim1
1Department of Biomedical Engineering, College of Health Science, Yonsei University, Wonju 220-710, 2Brain Korea 21 Projet for Medical
Science and Brain Research Institute, 3Department of Neurosurgery, Yonsei University College of Medicine, Seoul 120-752, 4Department
of Physiology, College of Medicine, Hallym University, Chuncheon 200-702, Korea
Deep brain stimulation (DBS) of the subthalamic nucleus (STN) has been widely used as a treatment
for the movement disturbances caused by Parkinson’s disease (PD). Despite successful application of
DBS, its mechanism of therapeutic effect is not clearly understood. Because PD results from the
degeneration of dopamine neurons that affect the basal ganglia (BG) network, investigation of neuronal
responses of BG neurons during STN DBS can provide informative insights for the understanding of
the mechanism of therapeutic effect. However, it is difficult to observe neuronal activity during DBS
because of large stimulation artifacts. Here, we report the observation of neuronal activities of the
globus pallidus (GP) in normal and PD model rats during electrical stimulation of the STN. A custom
artifact removal technique was devised to enable monitoring of neural activity during stimulation. We
investigated how GP neurons responded to STN stimulation at various stimulation frequencies (10,
50, 90 and 130 Hz). It was observed that activities of GP neurons were modulated by stimulation
frequency of the STN and significantly inhibited by high frequency stimulation above 50 Hz. These
findings suggest that GP neuronal activity is effectively modulated by STN stimulation and strongly
dependent on the frequency of stimulation.
Key Words: Artifact removal, Deep brain stimulation, Globus pallidus, Parkinson’s disease, Subthalamic
Parkinson’s disease (PD) is a neurodegenerative disorder
that causes considerable movement disability. It is well
known that motor symptoms of PD are caused by the death
of dopamine-generating cells in the pars compacta region
of the substantia nigra. To reverse parkinsonian symptoms,
pharmacological therapy using levodopa (L-dopa) has been
used, as L-dopa is converted into dopamine in the brain
. However, drugs eventually become ineffective due to
the gradual loss of dopaminergic neurons as the disease
progresses. Surgical treatments, such as lesioning specific
parts of the brain and deep brain stimulation (DBS), have
been used as alternatives.
The stimulation of the STN by electrical pulses has now
been accepted as a practical treatment for PD-related move-
ment disorders, including tremor and dystonia [2-8].
Several reports on neural activities evoked by DBS can be
found in the literature [9-13], yet the mechanism of DBS
is not clearly understood. Contradictory results exist on the
effect of DBS; it is unclear whether the STN DBS produces
an inhibitory or excitatory effect. For example, Benabid et
al. reported that high frequency DBS can mimic the effects
of lesioning [14,15]. Benazzouz et al. showed that STN stim-
ulation can cause inhibition of the STN excitatory output
[9,10]. However, Vitek et al. reported that lesions and DBS
could produce the opposite effects . In addition, Windels
et al. demonstrated that high frequency stimulation of STN
could increase firing of STN neurons . These contra-
dictions may be due to differences in the details of stim-
ulation parameters or differences in experimental protocols.
However, it is more likely that STN DBS could induce more
complex responses in the basal ganglia (BG). Shi et al. re-
ported that BG neurons of GP and substantia nigra pars
reticulate (SNr) showed both inhibitory and excitatory re-
sponses while behaviorally effective DBS was delivered to
the STN. These results indicate that the therapeutic effect
300 SB Ryu, et al
Fig. 1. Protocol for artifact removal with a curve-fitting method. (A) Single unit activities are covered by large stimulus artifact waveforms.
(B) An example of one artifact waveform at a stimulation frequency of 10 Hz. Several spikes were mixed on the fluctuation generated
by an artifact waveform (indicated by arrows). (C) The artifact waveform is divided into several segments and each segment is curve-fitted
with 2nd to 4th order polynomial. Every curve-fitted waveform is gathered together to construct a template for artifact subtraction. (D)
Spike waveforms were clearly revealed after artifact subtraction (the template is subtracted from the original waveforms). (E) Every artifact
waveform is subtracted by templates, which are curve-fitted from their own waveform. After this procedure is repeated, an artifact-removed
signal is obtained.
of DBS might not be caused by a simple inhibition or ex-
citation of the BG . Therefore, investigations of neuro-
nal activity in relevant areas in the BG as well as the
stimulated area are essential to clarify DBS mechanism.
Electrical stimulation produces large artifact waveforms
whenever neural recordings are performed simultaneously
with electrical stimulation. The artifacts significantly oc-
clude or completely block neural spike waveforms. To re-
solve this issue, several artifact removal techniques have
been devised [19-23]. Most artifact removal techniques em-
ploy a template subtraction method that attempts to re-
construct spike waveforms by subtracting a mean artifact
waveform from each recorded artifact waveform. These
techniques could generate residual artifacts due to the vari-
ability in the stimulation onset time or the variation in arti-
fact waveforms. To alleviate this problem, we developed a
custom artifact removal technique using a curve-fitting me-
thod to minimize possible residual artifacts by generating
a template from a single artifact waveform. Simulations
that test spike detection and classification from artifact-re-
moved signals validated the developed artifact removal
In this study, using the custom artifact removal techni-
que, we aimed to observe changes of neuronal activities in
the GP of normal and Parkinson’s disease model rats while
electrical stimulation was applied to the STN. Because it
is known that the therapeutic effects of DBS are strongly
dependent on the stimulation frequency, we investigated
how the frequency of STN stimulation affects neuronal ac-
tivities in the GP.
A total of 5 adult male Sprague-Dawley rats (200∼250
g) were used. Rats were divided into two groups as follows:
(i) a normal control group containing 2 rats without lesions
and (ii) a PD model group containing 3 rats with a medial
forebrain bundle (MFB) lesion induced by 6-hydroxydop-
amine (6-OHDA). Rats with MFB lesion has been proposed
as a good model of the akinesia associated with PD because
it induced deficits in forepaw adjusting steps in rats .
To explain briefly, rats were anesthetized with a mixture
of ketamine (75 mg/kg), acepromazine (0.75 mg/kg) and xy-
lazine (4 mg/kg), and mounted on a stereotaxic apparatus.
6-OHDA hydrobromide (8 μg free base in 0.2% ascorbic
acid, Sigma, St. Louis, MO) was injected unilaterally into
the MFB according to the following stereotaxic coordinates:
AP －4.4 mm, ML 1.2 mm relative to bregma, and DV －7.5
mm from the dura mater. In 6-OHDA injection, we made
the injection needles using stainless steel pipe (in 200 μm
diameter). The needles were sanitizes with E.O. gas and
connected with Hamilton microsyringe (80330, Hamilton
company, Reno, NV, U.S.A). The solution was injected at
a rate of 0.5 μl/min using a cannula and the injection was
pushed by hand. A polyethylene tube was used to connect
the cannula and microsyringe. To prevent noradrenergic
neurons from being destroyed, desipramine (12.5 mg/kg, in-
traperitoneally) was administered 30 min prior to the
6-OHDA infusion. Electrophysiological recordings were per-
formed 5 weeks after the 6-OHDA lesion.
Extracellular recording and electrical stimulation
Extracellular recordings of single unit activities from
rats were performed under anesthesia with a mixture of
Neuronal Responses in GP during STN DBS 301
Fig. 2. Comparison of two artifact removal methods. (A) 100 overlapped artifact waveforms (thick gray trace: mean waveform) recorded
with a stimulation frequency of 10 Hz. (B) Artifact-removed waveform after the mean waveform subtraction. Residual artifacts and
fluctuation remained (indicated by arrows). (C) Artifact-removed waveform after the template subtraction with the curve-fitting method.
Residual artifacts that are shown in Fig. 2 (B) were not generated using the curve-fitting method.
Fig. 3. Simulations for the validation of the effectiveness of the artifact removal technique with the curve-fitting method. (A, B) Two
different shapes of spikes were randomly synthesized to the noise (first 20 seconds) and to the noise and artifact waveforms (second 20
seconds). In this figure, stimulation frequency of the artifacts was 90 Hz (top: before artifact removal, bottom: after artifact removal).
(C) Spikes are detected after artifact removal and classified by principal component analysis (PCA). Spikes detected before and after artifact
subtractions have similar spike waveforms and can be reliably classified into two groups (In the box: spikes waveforms sorted by PCA
analysis. Black waveforms: spikes detected from the first 20 seconds. Grey waveforms: spikes detected from the second 20 seconds after
artifacts were removed).
Zoletil 50 (40 mg/kg) and xylazine (9 mg/kg). A micro-
electrode (A-M systems, Tungsten, diameter: 0.005 in, AC
impedance 5 MΩ, product no. 573400) was used for the
recording. The microelectrode was stereotaxically guided
through a burr hole drilled in the skull at the target coor-
dinates (GP, AP, －0.92 mm; ML, 3 mm; DV, 6.0∼6.6 mm).
A charge-balanced biphasic current pulse train was gen-
erated by a stimulator (A-M systems, isolated pulse stim-
ulator, model 2100) and applied to the subthalamic nucleus
(STN, AP, －3.7 mm; ML, 2.5 mm; DV, 7.4∼7.8 mm)
through a stimulation electrode (A-M systems, Tungsten,
diameter: 0.010 in, tip exposure: 1 mm, product no.563410)
with amplitudes of 100 and 200 μA, and a pulse duration
of 60 μs. Stimulation frequency was varied from 10 to 130
Hz (10, 50, 90 and 130 Hz). Electrical signals were ampli-
fied using an amplifier (A-M systems, microelectrode AC
amplifier, model 1800). The signals were stored on a PC
equipped with Spike 2 software (version 2.18, Cambridge
Electronic De sign, UK) and analyzed by using Matlab
302SB Ryu, et al
Fig. 4. Inhibition of neuronal activity of GP neurons by STN stimulation. (A) Inhibition of neuronal activity followed electrical stimulation.
Stimulation of a larger amplitude induced longer inhibition (top: 100 μA, bottom: 200 μA). (B, C) A post-stimulus time histogram (PSTH,
time bin: 10 ms) and raster plot were constructed from 50 repetitive stimulations. The duration of the inhibition was approximately 100
ms and 200 ms for pulse amplitude 100 μA and 200 μA, respectively.
Spike detection and sorting accuracy
after artifact removal (%)
10 Hz50 Hz 90 Hz 130 Hz
Table 1. Simulation results. Spike detection and sorting accu-
racy after artifact removal
Artifact removal and data analysis
Fig. 1A shows a typical waveform recorded from the GP
during electrical stimulation applied to the STN. Although
it is possible to observe spikes during the fluctuation of the
artifact waveform by adjusting the time and amplitude
scale, it is not easy to detect spikes by an amplitude thresh-
old (Fig. 1B). To remove the fluctuation and detect spikes,
a template for the artifact subtraction was constructed by
a curve-fitting method. As shown in Fig. 1C, the artifact
waveform of the trace in Fig. 1B was divided into several
segments and each segment is curve-fitted with a 2nd to
4th order polynomial. The length of each segment is ad-
justed according to the shape of the segment; usually a 50∼
100 sample length (2∼4 ms) was appropriate in the present
study. The order of the polynomial was decided not to
curve-fit the spike waveform in the artifact. After all the
segments were curve-fitted, the segments were gathered to
construct a template (Fig. 1C). For our experimental data,
a period of 2 ms from the stimulation onset was flattened
because the artifact waveform is saturated, and spikes can-
not be detected in this period. For the final step, the con-
structed template was subtracted from the original artifact
waveform. As a result, the artifact waveform and the fluctu-
ation were effectively removed and spikes were clearly re-
vealed (Fig. 1D and Fig. 2). This procedure was repeated
for all the artifact waveform and finally, the artifact-free
neural signal could be obtained (Fig. 1E). After the artifacts
were removed, spikes were detected and sorted by the prin-
cipal component analysis (PCA) as shown in Fig. 3C. To
quantitatively assess changes in neuronal activities with
electrical stimulation, firing rate before and during the elec-
trical stimulation was calculated and compared.
Simulation of artifact removal
To validate the effectiveness of our custom artifact re-
moval, a simulation was performed to detect and classify
spikes under conditions when spikes were mixed with arti-
fact waveforms. For the simulation, stimulation artifact
waveforms (of various stimulation frequencies: 10, 50, 90
and 130 Hz) and two different shapes of spike waveforms
(two units) were extracted from an experimental dataset.
The simulation consisted of control and test sessions. The
control session was 20 seconds long, and two different spike
waveforms were randomly synthesized on the noise (the to-
tal number of spikes is 800 which means that 400 spikes
per unit were used in the simulation). The test session was
also 20 seconds long, and two different spike waveforms
were randomly synthesized on the noise signal and on the
stimulation artifact waveforms, avoiding each stimulation
onset (Fig. 3A, top). After the signal for the simulation was
generated, artifact waveforms were removed by the artifact
removal technique (Fig. 3A and 3B). Spikes were detected
Neuronal Responses in GP during STN DBS303
Cell 1Cell 2
*Paired-sample t-test. NS, not significant.
Table 2. Mean firing rate of GP cells before and during STN stimulation
Fig. 5. Waveforms recorded from GP neurons of Parkinson’s disease model rats. (A) Electrical stimulations were applied to the STN with
various stimulation frequencies (from the top: 10, 50, 90 and 130 Hz). (B) Artifact-removed waveforms. Artifact-removed periods are indicated
with boxes of dotted lines. (C) Spike waveforms detected from each waveform. Two units were detected and sorted. Spikes with amplitude
larger than another unit showed inhibition of firing rate when high frequency stimulation was applied to the STN.
by threshold and sorted by their waveforms using the prin-
cipal component analysis (Fig. 3C). Although the ampli-
tudes of some spikes were diminished in size, they were
reliably classified into two groups. The simulation was re-
peated 10 times for each stimulation frequency and the ac-
curacy was calculated by checking whether the timestamps
of detected spike trains were correct. As shown in Table
1, the average accuracy of detected spikes was approx-
imately 80∼90%, which means that over 80% of spikes
were successfully detected from artifact-contaminated wa-
Responses of GP neurons during STN stimulation
Fig. 4 shows response characteristics of a GP neuron in
a PD model rat to electrical stimulation of the STN. As
shown in Fig. 4A, inhibitory periods followed stimulation
pulses and stronger stimulation induced longer inhibition
(top: 100 μA, bottom: 200 μA). When an electrical pulse
was applied repeatedly 50 times with a 5 s interval, in-
hibition was consistently observed in the raster plot (Fig.
4C). From the poststimulus time histogram (PSTH) in Fig.
4B, it was found that the duration of the inhibitory period
was approximately 100 and 200 ms when the pulse ampli-
tude was 100 μA and 200 μA, respectively.
Fig. 5A demonstrates neural activities of GP neurons in
a PD model rat, recorded with stimulation artifacts. Be-
cause it is impossible to observe spikes during STN stim-
ulation due to these stimulation artifacts, the artifacts were
eliminated with the artifact removal technique. After arti-
fact removal, spikes were detected and classified into two
units (Fig. 5B and 5C). As seen in Fig. 5B, spikes with large
amplitudes were modulated by high-frequency STN stim-
ulation and significantly suppressed when stimulation fre-
quency was over 50 Hz (This result is demonstrated in
Table 2, PD Cell 1).
To quantitatively assess changes in the firing rate due
to STN stimulation, we calculated and compared relative
changes in the firing rate between two periods- the 10 s
preceding stimulation and during the 10 s stimulation. As
304SB Ryu, et al
Fig. 6. Changes in the firing rate according to the stimulation frequency. Changes of the mean firing rate during electrical stimulation
were compared to the mean firing rate 10 s before the stimulation to obtain quantitative values (relative change, %). Both in (A) normal
and (B) PD model rat, significant inhibition of neuronal firing was observed at stimulation frequencies higher than 50 Hz.
shown in Fig. 6, the relative change in the firing rate of
GP cells decreased according to the stimulation frequency.
In order to check whether its change was significant, stat-
istical test was performed for every cell in PD model rats
(Table 2). It was found that the firing rate was significantly
decreased when the STN was electrically stimulated with
frequency higher than 50 Hz.
In the present study, we observed neuronal activities in
the GP of normal and PD model rats responding to STN
stimulation. According to the traditional model of the BG-
thalamocortical network , the pathophysiological mech-
anisms underlying movement disorders of PD are asso-
ciated with abnormal outflows within the BG. It is regarded
that neuronal responses within the BG directly reflect the
effect of STN stimulation. Therefore, characterizing BG
neuron responses can help us understand the working me-
chanism of DBS. We specifically focused on observing GP
neuronal activities during STN stimulation using a custom
artifact removal technique and demonstrated how re-
sponses changed with the frequency of STN stimulation.
Previous studies mainly investigated neuronal responses
that occurred after the termination of electrical stimulation
with the assumption that neuronal activities following the
stimulation may reflect activities that occurred during
stimulation [9,10]. However, Montgomery et al.  demon-
strated that increased neuronal activities during electrical
stimulation were reduced after the cessation of the stim-
ulation, implying that neuronal activities following elec-
trical stimulation may show an effect opposite to that which
occurs during the stimulation. Accordingly, the observation
of neuronal activities during electrical stimulation is re-
garded as an important tool for the study of the mechanism
Single pulse stimulation of the STN induced inhibitory
responses in the GP (Fig. 4). However, the inhibitory effect
was very brief and recovered quickly and it is not expected
that a single pulse can generate clinical effects. We ob-
served neural activities in the GP responding to a pulse-
train stimulation of the STN under various stimulation
frequencies. GP activity was not significantly suppressed
in PD model rats when the stimulation frequency was as
low as 10 Hz. However, GP activity was remarkably in-
hibited when the stimulation frequency increased to be
greater than 50 Hz (Fig. 6, Table 2). This result supports
the indirect inhibition of pathological activity hypothesis
, which suggests that high-frequency stimulation is ef-
fective because there is less opportunity for abnormal neu-
ronal activity to return to the baseline pathological activity
due to the short inter-stimulus pulse intervals. A stim-
ulation frequency of 50 Hz may be inconsistent with gen-
erally accepted stimulation parameters such as 130 Hz.
However, as shown in the results of the single pulse stim-
ulation, neuronal activities are affected not only by stim-
ulation frequency but also by stimulation intensity. Thus,
for a direct comparison of the results on stimulation fre-
quency, other conditions such as stimulation intensity,
pulse width, stimulation period and subjects group should
be matched. Nevertheless, our results are in good agree-
ment with the results of Limousin et al. , which demon-
strated that a beneficial effect of STN stimulation on par-
kinsonian motor symptoms was dependent on stimulation
frequency and the maximum effect occurs over 50 Hz.
Therefore, we hypothesize that STN stimulation could be
effective for the alleviation of parkinsonian symptoms when
the stimulation parameter is strong enough to inhibit neu-
We used an artifact removal technique devised from a
curve-fitting method as an essential tool in this study.
Stimulation artifact is a major problem in the study of neu-
ronal activities responding to the high-frequency electrical
stimulation as artifact waveforms not only occupy most of
the recording time but also occlude neural signals. Several
studies employed a template subtraction method that sub-
tracts a mean artifact waveform (template waveform) from
each artifact waveform and showed that is effective for re-
vealing spikes occluded by artifact waveforms. However,
when we tested the template subtraction method using a
mean artifact waveform, there were unintended residual
artifacts and fluctuations left after the subtraction (Fig.
2B). Although residual artifacts due to the variation of the
Neuronal Responses in GP during STN DBS305
stimulation onset can be removed by flattening the short
segment before and after the stimulation onset, residual
fluctuations cannot be easily flattened because the length
of the fluctuation may be long enough to contain spikes.
This may be the reason why Hashimoto et al.  empha-
sized that it is critically important to construct an accurate
template for this method. Another possible problem is hav-
ing evoked spikes with a fixed latency. If spikes are evoked
with a fixed latency responding to the stimulation, a tem-
plate constructed by averaging artifact waveforms may con-
tain the spike shape in the template and can eliminate each
evoked spike by the template subtraction. Our artifact re-
moval technique was devised to avoid the problems of the
mean waveform subtraction. Because the template is gen-
erated from a single artifact waveform by curve-fitting each
waveform, it can minimize the variance between the tem-
plate and each artifact waveform and, therefore, prevent
the residual artifacts. This was most effective when the
stimulation frequency was low so that the length of an arti-
fact waveform was several tens of milliseconds. In addition,
it could reduce the concern for the elimination of spikes
with a fixed latency, as the template is generated to fit each
artifact waveform. Moreover, this method can be applied
not only to sequential pulse-trains but also to single pulses
because it is not necessary to collect many artifact wave-
forms to construct an average waveform template.
Here, we observed that neuronal activities of the GP are
inhibited by high frequency STN stimulation. However, it
is still necessary to investigate neuronal responses of other
BG neurons as well as the cortex to understand the me-
chanism. Therefore, in future studies, simultaneous multi-
channel recording from the BG and the cortex should be
performed during electrical stimulation. Observation of sin-
gle unit activity as well as local field potentials from vari-
ous sites within the BG will also provide valuable in-
formation about the mechanism of DBS.
This study was supported by the Yonsei University
Research Fund of 2013 and a grant from the Industrial
Source Technology Development Program of the Ministry
of Knowledge Economy (MKE) of Korea (no. 10033812) and
the grant from the Center for Integrated Smart Sensors
funded by the Ministry of Science, ICT & Future Planning
as Global Frontier Project (CISS-2011-0038167).
1. The National Collaborating Centre for Chronic Conditions,
National Clinical Guideline for Diagnosis and Management in
Primary and Secondary Care. London: Royal College of Phy-
sicians (UK); 2006.
2. Benabid AL, Pollak P, Gervason C, Hoffmann D, Gao DM,
Hommel M, Perret JE, de Rougemont J. Long-term suppression
of tremor by chronic stimulation of the ventral intermediate
thalamic nucleus. Lancet. 1991;337:403-406.
3. Koller W, Pahwa R, Busenbark K, Hubble J, Wilkinson S, Lang
A, Tuite P, Sime E, Lazano A, Hauser R, Malapira T, Smith
D, Tarsy D, Miyawaki E, Norregaard T, Kormos T, Olanow CW.
High-frequency unilateral thalamic stimulation in the treat-
ment of essential and parkinsonian tremor. Ann Neurol.
4. Kumar R, Lozano AM, Sime E, Halket E, Lang AE. Com-
parative effects of unilateral and bilateral subthalamic nucleus
deep brain stimulation. Neurology. 1999;53:561-566.
5. Limousin P, Pollak P, Benazzouz A, Hoffmann D, Le Bas JF,
Broussolle E, Perret JE, Benabid AL. Effect of parkinsonian
signs and symptoms of bilateral subthalamic nucleus sti-
mulation. Lancet. 1995;345:91-95.
6. Rodriguez-Oroz MC, Obeso JA, Lang AE, Houeto JL, Pollak
P, Rehncrona S, Kulisevsky J, Albanese A, Volkmann J, Hariz
MI, Quinn NP, Speelman JD, Guridi J, Zamarbide I, Gironell
A, Molet J, Pascual-Sedano B, Pidoux B, Bonnet AM, Agid Y,
Xie J, Benabid AL, Lozano AM, Saint-Cyr J, Romito L,
Contarino MF, Scerrati M, Fraix V, Van Blercom N. Bilateral
deep brain stimulation in Parkinson's disease: a multicentre
study with 4 years follow-up. Brain. 2005;128:2240-2249.
7. Siegfried J, Lippitz B. Bilateral chronic electrostimulation of
ventroposterolateral pallidum: a new therapeutic approach for
alleviating all parkinsonian symptoms. Neurosurgery. 1994;35:
8. Yianni J, Bain P, Giladi N, Auca M, Gregory R, Joint C, Nandi
D, Stein J, Scott R, Aziz T. Globus pallidus internus deep brain
stimulation for dystonic conditions: a prospective audit. Mov
9. Benazzouz A, Gao DM, Ni ZG, Piallat B, Bouali-Benazzouz R,
Benabid AL. Effect of high-frequency stimulation of the
subthalamic nucleus on the neuronal activities of the sub-
stantia nigra pars reticulata and ventrolateral nucleus of the
thalamus in the rat. Neuroscience. 2000;99:289-295.
10. Benazzouz A, Piallat B, Pollak P, Benabid AL. Responses of
substantia nigra pars reticulata and globus pallidus complex
to high frequency stimulation of the subthalamic nucleus in
rats: electrophysiological data. Neurosci Lett. 1995;189:77-80.
11. Dostrovsky JO, Lozano AM. Mechanisms of deep brain
stimulation. Mov Disord. 2002;17 Suppl 3:S63-68.
12. Maurice N, Thierry AM, Glowinski J, Deniau JM. Spontaneous
and evoked activity of substantia nigra pars reticulata neurons
during high-frequency stimulation of the subthalamic nucleus.
J Neurosci. 2003;23:9929-9936.
13. Tai CH, Boraud T, Bezard E, Bioulac B, Gross C, Benazzouz
A. Electrophysiological and metabolic evidence that high-
frequency stimulation of the subthalamic nucleus bridles
neuronal activity in the subthalamic nucleus and the
substantia nigra reticulata. FASEB J. 2003;17:1820-1830.
14. Benabid AL, Pollak P, Gao D, Hoffmann D, Limousin P, Gay
E, Payen I, Benazzouz A. Chronic electrical stimulation of the
ventralis intermedius nucleus of the thalamus as a treatment
of movement disorders. J Neurosurg. 1996;84:203-214.
15. Benabid AL, Pollak P, Gross C, Hoffmann D, Benazzouz A, Gao
DM, Laurent A, Gentil M, Perret J. Acute and long-term effects
of subthalamic nucleus stimulation in Parkinson's disease.
Stereotact Funct Neurosurg. 1994;62:76-84.
16. Vitek JL, Hashimoto T, Peoples J, DeLong MR, Bakay RA.
Acute stimulation in the external segment of the globus
pallidus improves parkinsonian motor signs. Mov Disord. 2004;
17. Windels F, Bruet N, Poupard A, Urbain N, Chouvet G,
Feuerstein C, Savasta M. Effects of high frequency stimulation
of subthalamic nucleus on extracellular glutamate and GABA
in substantia nigra and globus pallidus in the normal rat. Eur
J Neurosci. 2000;12:4141-4146.
18. Shi LH, Luo F, Woodward DJ, Chang JY. Basal ganglia neural
responses during behaviorally effective deep brain stimulation
of the subthalamic nucleus in rats performing a treadmill
locomotion test. Synapse. 2006;59:445-457.
19. Harding GW. A method for eliminating the stimulus artifact
from digital recordings of the direct cortical response. Comput
Biomed Res. 1991;24:183-195.
20. Hashimoto T, Elder CM, Vitek JL. A template subtraction
method for stimulus artifact removal in high-frequency deep
brain stimulation. J Neurosci Methods. 2002;113:181-186.
21. Heffer LF, Fallon JB. A novel stimulus artifact removal
technique for high-rate electrical stimulation. J Neurosci
306 SB Ryu, et al Download full-text
22. O'Keeffe DT, Lyons GM, Donnelly AE, Byrne CA. Stimulus
artifact removal using a software-based two-stage peak
detection algorithm. J Neurosci Methods. 2001;109:137-145.
23. Wichmann T. A digital averaging method for removal of
stimulus artifacts in neurophysiologic experiments. J Neurosci
24. Chang JW, Wachtel SR, Young D, Kang UJ. Biochemical and
anatomical characterization of forepaw adjusting steps in rat
models of Parkinson's disease: studies on medial forebrain
bundle and striatal lesions. Neuroscience. 1999;88:617-628.
25. DeLong MR. Primate models of movement disorders of basal
ganglia origin. Trends Neurosci. 1990;13:281-285.
26. Montgomery EB Jr. Effect of subthalamic nucleus stimulation
patterns on motor performance in Parkinson's disease.
Parkinsonism Relat Disord. 2005;11:167-171.
27. Montgomery EB Jr, Gale JT. Mechanisms of action of deep
brain stimulation (DBS). Neurosci Biobehav Rev. 2008;32: