Post on 02-Jan-2017
ORIGINAL ARTICLE
Evoked potential and behavioral outcomes for experimentalautoimmune encephalomyelitis in Lewis rats
Angelo H. All • Gracee Agrawal • Piotr Walczak •
Anil Maybhate • Jeff W. M. Bulte • Douglas A. Kerr
Received: 20 November 2008 / Accepted: 23 April 2010
� Springer-Verlag 2010
Abstract A reliable outcome measurement is needed to
assess the effects of experimental lesions in the rat spinal
cord as well as to assess the benefits of therapies designed
to modulate them. The Basso, Beattie, and Bresnahan
(BBB) behavioral scores can be indicative of the func-
tionality in motor pathways. However, since lesions are
often induced in the more accessible dorsal parts associated
with the sensory pathways, the BBB scores may not be
ideal measure of the disability. We propose somatosensory
evoked potential (SEP) as a complementary measure to
assess the integrity of sensory pathways. We used the focal
experimental autoimmune encephalomyelitis (EAE)
model, in which focal demyelinating lesions were induced
by injecting cytokine-ethidium bromide into dorsal white
matter after MOG-IFA immunization. Both the SEP and
BBB measures reflected injury; however, the SEP was
uniformly and consistently altered after the injury whereas
the BBB varied widely. The results suggest that the SEP
measures are more sensitive and reliable markers of focal
spinal cord demyelination compared to the behavioral
measures like the BBB score.
Keywords Somatosensory evoked potential �BBB score � Rat model � Focal experimental autoimmune
encephalomyelitis (EAE) � Cytokine-ethidium bromide �MOG-IFA immunization
Introduction
Multiple sclerosis (MS) affects about 2.5 million people
worldwide [1]. It is an autoimmune disease in which the
immune system of the body destroys the oligodendrocytes
responsible for creating and maintaining the myelin sheath
around the nerve fibers. This results in thinning or complete
loss of myelin in the central nervous system (CNS): brain,
spinal cord, and optic nerves. This often progresses to phys-
ical and cognitive disability in the affected [2, 3]. There is no
cure for MS. To develop better therapies for MS, researchers
must continue to develop and validate animal models for MS.
However, it is clear that the outcome measures for CNS injury
in animals are lacking and unreliable [4].
Experimental autoimmune encephalomyelitis (EAE) is a
rodent model of multiple sclerosis. In one version of EAE,
a focal demyelinating lesion is induced in the spinal cord
A. H. All � G. Agrawal � A. Maybhate � J. W. M. Bulte
Department of Biomedical Engineering, Johns Hopkins
University School of Medicine, Baltimore, MD, USA
A. H. All (&) � D. A. Kerr
Department of Neurology, Johns Hopkins
University School of Medicine, 720 Rutland Avenue,
Traylor 701C, Baltimore, MD 21205, USA
e-mail: hmn@jhu.edu
P. Walczak � J. W. M. Bulte
Department of Radiology and Radiological Science,
Johns Hopkins University School of Medicine,
Baltimore, MD, USA
P. Walczak � J. W. M. Bulte � D. A. Kerr
Institute for Cell Engineering, Johns Hopkins
School of Medicine, Baltimore, MD, USA
J. W. M. Bulte
Department of Chemical and Biomolecular Engineering,
Johns Hopkins University School of Medicine,
Baltimore, MD, USA
D. A. Kerr
Department of Molecular Microbiology and Immunology,
Johns Hopkins University School of Public Health,
Baltimore, MD, USA
123
Neurol Sci
DOI 10.1007/s10072-010-0329-y
by administering inflammatory factors directly into the
spinal cord of MOG-IFA immunized rats [5]. This focal
EAE model is analogous to the human paralyzing disorder
transverse myelitis (TM), which often arises idiopathically
or as part of multiple sclerosis. This is believed to be a
relevant model to study the neurological manifestations of
a focal inflammatory and demyelinating lesion [6]. There is
a need for objective reliable tools to assess the outcomes of
the lesion in the spinal cord, and also to assess the efficacy
of the new therapeutic techniques.
The behavioral measure used to assess the locomotor
functions after injury to spinal cord is the Basso, Beattie,
Bresnahan (BBB) Score [7, 8]. The BBB scale ranges
from 0 (no hindlimb function) to 21 (perfect hindlimb
function) using hindlimb movement, paw placement and
coordination as key determining factors [7, 8]. Other fine
movements, that the scale is sensitive to, include trunk
movement, stepping, toe clearance and tail position [7, 8].
BBB is a well established and easily executable tech-
nique. However, being a qualitative test relying on a few
minutes of subjective observations, it often has limited
sensitivity. Moreover it does not correspond directly to the
entire spectrum of possible outcomes [9–11], since lesion/
injury in most research studies is often induced in the
more accessible dorsal (sensory) pathways of the spinal
cord.
Somatosensory evoked potential (SEP) is an objective
and reliable electrophysiological measurement, which
reflects the status of the sensory pathways through the
spinal cord. It is the electrical response of the nervous
system to a sensory stimulus, recorded from the somato-
sensory cortex in the brain [12]. SEP is often used for
clinical studies as a tool to provide information regarding
functional integrity of the spinal cord [13–16].
In the present study, we used behavioral and electro-
physiological assessments to investigate the outcomes of
the focal EAE model on rodents. Our goal was to con-
currently study the outcomes of SEP and BBB measures
after induction of a focal inflammatory and demyelinating
lesion to the spinal cord.
Materials and methods
Animals
The experiment used adult female Lewis rats with an
average body weight of 200–220 gm, which were indi-
vidually housed with free access to food and water. All
experimental procedures were in accordance with the
guidance provided in the rodent survival surgery manual
and were approved by the Institutional Animal Care and
Use Committee at the Johns Hopkins University.
Anesthesia
For anesthesia, the rat was held in a transparent chamber
with 3% isoflurane gas and room air flow until the onset of
drowsiness. Its mouth and nose was then placed within an
anesthesia mask (with a well-fitting rodent size dia-
phragm), which was connected to a C-Pram circuit
designed to deliver and evacuate the gas through one tube.
A mixed flow of 1% isoflurane, 80% oxygen and room air
was delivered to the mask at the rate of 1.0 L/min. Rats
continued spontaneous breathing and the anesthesia depth
level was maintained unchanged throughout the entire
experiment. Rats were also placed on a homoeothermic
blanket to maintain their body temperature at 37.5 ± 0.5�C
throughout the entire experiment.
SEP electrode implantation
After anesthetizing the rats, an incision was made on the
rat’s midline and the cranium bone was cleaned. This was
followed by implantation of five transcranial screw elec-
trodes (E363/20, Plastics One Inc.) on the somatosensory
cortex as shown in Fig. 1a. The forelimb recording elec-
trodes were implanted 0.2 mm posterior to the bregma and
3.8 mm laterally from the bregma, and the hindlimb
recording electrodes were implanted 2.5 mm posterior to
the bregma and 2.8 mm laterally from the bregma on each
hemisphere. A fifth electrode on the right frontal bone,
situated 2 mm from both the sagittal and coronal sutures,
served as the intracranial reference. The electrodes made
very light contact with the dura mater without compressing
the dura or brain structures.
Focal lesion induction
Recombinant myelin oligodendrocyte glycoprotein (MOG)
corresponding to the N-terminal sequence of rat MOG
amino acids 1–125 (MOG1–125) was emulsified in
incomplete Freund’s adjuvant (Sigma-Aldrich) as 1:1
mixture (Imject IFA; Pierce). 100 ll = 50 lg of this
emulsified MOG1–125 (from Dr. Sha Mi, Biogen-Idec,
Cambridge, MA, USA) was injected subcutaneously near
the base of the tail of each rat at two contralateral sites
(50 ll per area to minimize irritation to the skin).
The rats were subjected to T9 laminectomy and
(2 9 2 ll) cytokines (250 ng of TNF-a, 150U of INF-c and
40 ng of IL-6) and ethidium bromide (1 lg) were injected
into the dorsal white matter using a Hamilton needle (31G).
Behavioral assessment
The motor function was assessed using the BBB locomotor
rating scale before injury and days 4 and 7 post-injury. This
Neurol Sci
123
scoring scale is sensitive to rat hindlimb movements, trunk
position and stability, coordination, stepping and paw
placement, and tail position; and the details of the scoring
scheme can be found in [7, 8]. The increasing BBB scores
correspond to an increasing level of functional recovery,
and can be categorized into three phases of recovery: early,
intermediate, and late.
Rats were placed individually in a 90 cm plastic open
field. Two examiners observed and scored each rats for
exactly 4 min, giving a separate score for each hindlimb.
The final given score for each limb was the minimum
between the two observers. The rats were scored in a range
0–21, 0 signifying no hindlimb movement and 21 signi-
fying perfect hindlimb movement with extensive move-
ment of hindlimbs, consistent plantar stepping, consistent
toe clearance, no rotation in stepping, and tail consistently
up [7, 8].
Multilimb SEP recording and analysis
Thirty-minute SEP recordings were conducted before
injury and days 4 and 7 post-injury. Rats were anesthetized
during each recording as described before, and the anes-
thesia level and body temperature were maintained.
The median and tibial nerves of both left and right limbs
were electrically stimulated using subcutaneous needle
electrodes (Safelead F-E3-48, Grass-Telefactor, West
Warwick, RI, USA) and an isolated constant current
stimulator (DS3, Digitimer Ltd., Hertfordshire, England),
without making any direct contact with the nerve bundle.
The stimulator was triggered using a custom intraoperative
neurological monitoring (INM) software (Infinite Bio-
medical Technologies, Baltimore, MD, USA), which was
also used to set the stimulation parameters [17]. The limbs
were stimulated by positive current pulses of 3.5 mA
(a)
(b)
10 msec
100
µV
10 msec
50
µV
yrujni-erpyrujni-erp
day4post-injury
day4post-injury
day7post-injury
day7post-injury
RightForelimb
LeftForelimb
RightHindlimb
LeftHindlimb
Fig. 1 a A schematic of the rat
skull with electrode placements
for SEP recording b SEP signal
at different experimental stages
Neurol Sci
123
magnitude, 200 ls duration, and 1 Hz frequency. An
advanced demultiplexer circuit was used to sequentially
stimulate each of the four limbs at a frequency of 0.25 Hz.
An optically isolated biopotential amplifier (Opti-Amp
8002, Intelligent Hearing Systems, Miami, FL, USA) was
used to record cortical somatosensory evoked potentials
from the transcranial electrodes with a gain of 30,000. The
analog signal was sampled at a rate of 5,000 Hz using an
optical data acquisition system with four input channels. The
SEP signals from each hemisphere, the stimulation pulse
signal, and the code for identifying the stimulated limb were
recorded on separate channels for offline data analysis.
Contralateral SEP recordings were used for analysis,
which was performed with Matlab 7.0. The signal-to-noise
ratio (SNR) was improved by ensemble averaging of 100
stimulus-locked sweeps, with the averaging window shifting
by 20 sweeps each time. The analysis of multilimb SEP
recordings was performed by using ratiometric analysis of
amplitude changes—hindlimb: forelimb peak-to-peak
amplitude ratio—between both the left and right pairs of
limbs [14]. This ratio is unaffected by systemic changes in
blood pressure or anesthesia, since signal changes due to
them are expected to be approximately equal for each limb.
Signal changes due to the lesion, however, can be detected
since injury at the thoracic level would cause the hindlimb
SEP signals to be modulated significantly more than their
forelimb counterparts. To account for inter-animal vari-
ability, the post-surgical SEP amplitude ratios were
expressed as a percentage of the baseline amplitude ratios of
each rat, and were referred to as normalized SEP amplitudes.
Statistical analysis
The commercially available software R (v2.7.2; The R
Foundation for Statistical Computing) was used for all the
statistical analysis, and p values of less than 0.01 were
considered statistically significant. Statistical analysis of
the SEP data was first performed by a repeated measures
analysis of variance (ANOVA) over three time points (pre-
injury, day 4 post-injury, day 7 post-injury), using the
normalized SEP amplitude as the input. The behavioral
outcomes were also analyzed in a similar way. The two
measures were compared using a two-way repeated mea-
sures analysis of variance (ANOVA) over the three time
points with the measure (SEP, BBB) as a main factor.
Results
Neuroelectrophysiological assessment
Figure 1 shows SEP signals from a sample rat for all the
limbs and all the study times (pre-injury, day 4 post-injury,
day 7 post-injury). As was expected, it can be seen that
there is no major change in the forelimb SEP signals after
injury; however, the hindlimb SEP signals degrade signif-
icantly from the baseline signals due to injury. This was
evident in both the left and right hind limb (LH and RH).
The pre-existing morphological differences in the left and
right side baseline SEP could be due to an asymmetric
placement of the electrodes described in the ‘‘Methods’’
section as shown in Fig. 1a. There are mainly two reasons
for the difference. One is related to the location of the
recording electrodes in skull. Although we have a precise
stereotaxic coordination for rat’s skull recording electrodes
implantation, not only the two LF and RF electrodes, as is
shown in Fig. 1a, could be off by a few microns or so but
also their individual distances to the reference electrodes
are different obviously. The second reason is related to the
positioning of the stimulus electrode. The distance from the
two pairs, left and right, stimulation electrodes, which are
inserted near tibial nerves through skin, are proximate.
Therefore, the two electrodes might induce slightly dif-
ferent stimulation in tibial nerves and invoke slightly dif-
ferent responses and consequently the baseline recordings
will not be identical from the two sides.
An inflammatory and demyelinating lesion manifests
itself as a significant reduction in the peak-to-peak ampli-
tude of SEP waveforms [18, 19]. The normalized SEP
amplitudes for both sides at three stages of the experiment
are shown in Fig. 2. All the rats demonstrated significant
injury as indicated by a large reduction in the SEP ampli-
tudes (C70% reduction) on day 4 after injury, although
some rats showed a marginal (statistically insignificant)
increase in SEP amplitudes on day 7. We have always
observed the worst EP signals on day 4 post-injury
recordings. This is related to acute phase of injury due to
surgery, specifically, incision of paravertebral muscles at
T6–12, laminectomy at T8–9 and the induction of intra-
parenchymal injury (EAE). All these could contribute to
the sharp artifactual reduction of SEP amplitudes imme-
diately after injury in the acute period (typically 3–4 days).
The wound healing process will take place during the first
week post-injury without any specific time line. Our EP
recording assessment is sensitive to this primary injury and
acute phase of injury and it is always reflected in our
recordings. Upon statistical comparisons (Student’s t test,
unpaired, one-tailed) we found that the difference in the
normalized SEP amplitudes across the different time points
was significant in both the left hindlimb (LH: p \ 0.0001)
and the right hindlimb (RH: p \ 0.0001).
Behavioral assessment
The early stages of recovery are characterized by little or
no hindlimb joint movement (BBB score 0–7), the
Neurol Sci
123
intermediate by bouts of uncoordinated stepping (8–13),
and the late by coordinated stepping with paw rotation, toe
dragging, and trunk instability (14–21) [7, 8]. The BBB
scores for both the hindlimbs over all the time points are
shown in Fig. 3, which illustrates that the rats’ BBB scores
are spread out into all the phases of recovery. The differ-
ence in BBB scores across the different time points was
also found to be significant in both the left hindlimb
(LH: p = 1.080e - 08) and right hindlimb (RH: p =
2.333e - 06). Again as in the SEP, the BBB scored
showed a marginal (statistically insignificant) improvement
from day 4 to day 7 which could be due to the main two
reasons explained above as well as abating of initial effects
of surgery as elaborated for SEP.
Electrophysiological and behavioral measures
Using two-way repeated measures ANOVA, we found a
significant difference between the two measures (SEP and
BBB) in both the left hindlimb (LH: p = 0.0011310) and
the right hindlimb (RH: p = 0.0009758). Also as expected,
there was a significant difference across the time points too
(LH: p = \2.2e - 16, RH: p = \2.2e - 16).
Histological verification of demyelination and axonal
damage
In order to verify the demyelination of the spinal cord and
the extent of axonal damage, rats were transcardially
pefused, spinal cord tissue collected and cryoprocessed for
histology. Thirty micrometers thick axial sections of spinal
cord were stained for hematoxylin and eosin to identify
regions of inflammatory and gliotoxic damage. Eriochrome
staining was used to visualize regions of myelin loss and
silver staining to delineate axonal density and extent of
axonal loss. We found that focal lesions were largely
confined to dorsal funiculus with only partial involvement
of surrounding gray matter (see Fig. 4). Lesions were
hypercellular (Fig. 4a) as compared to normal white matter
(Fig. 4b). Eriochrome staining demonstrated complete loss
0
20
40
60
80
100S
EP
N1P
2 A
mp
litu
de
Time point
Left Hindlimb(a)
0
20
40
60
80
100
pre-injury day4 post-injury day7 post-injury
pre-injury day4 post-injury day7 post-injury
SE
P N
1P2
Am
plit
ud
e
Time point
Right Hindlimb(b)
rat1rat2rat3rat4rat5rat6rat7rat8rat9rat10
rat1rat2rat3rat4rat5rat6rat7rat8rat9rat10
Fig. 2 Variations of normalized SEP amplitudes for 10 rats over
different time points for a left hindlimb, and b right hindlimb
0
3
6
9
12
15
18
21
pre-injury day4 post-injury day7 post-injury
pre-injury day4 post-injury day7 post-injury
BB
B S
core
Time point
Left Hindlimb(a)
0
3
6
9
12
15
18
21
BB
B S
core
Time point
Right Hindlimb(b)
rat1rat2rat3rat4rat5rat6rat7rat8rat9rat10
rat1rat2rat3rat4rat5rat6rat7rat8rat9rat10
Fig. 3 Variations of BBB score for 10 rats over different time points
for a left hindlimb, and b right hindlimb
Neurol Sci
123
of myelin in entire dorsal funiculus, which can be seen as
lack of blue color (Fig. 4c) abundant in unlesioned dorsal
funiculus (Fig. 4d). Silver staining for axons showed that
central part of dorsal funiculus had marked brightening
(Fig. 4e, arrows) and that represents some degree of axonal
loss from the center of the lesion. Normal axons are shown
in (Fig. 4f).
These results indicate that demyelination and some
axonal damage affect primarily dorsal sensory pathways.
Pathways participating in afferent motor conduction are
located in deep portion of the dorsal funiculus (in the
periphery of the lesion) as well as in the ventral portions of
the spinal cord, which was never affected by the lesion. For
these reasons we would expect to see consistent impair-
ment of sensory function and more variability in motor
deficits. The BBB scores are designed to assess the
motor behavior and reflects the functional status of the
motor pathways more directly (although it can indirectly
be affected by the functional status of the sensory path-
ways, too). Due to only a partial involvement of the
cortico-spinal tract our results with motor behavioral scores
showed more variability in the acute periods after injury
and later, compared to the more objective SEP amplitudes
which showed less variability and hence more reliability.
Conclusion and discussion
An implicit assumption in any therapeutic procedure is that
all injuries result into a similar functional decline in the
central nervous system. This is not always true and there-
fore we believe that the injuries to the sensory and motor
pathways must be distinguished to assess the extent of total
injury. Further, it should be noted that behavioral mea-
surements, by their nature, characterize mainly secondary
injury in motor pathways caused due to the lesion in
sensory pathways and may not be an effective way to
assess the injury [11].
In the present study, we concurrently studied the neu-
roelectrophysiology (SEP) with motor behavior (BBB)
outcome measurements in rats with focal injury. Both the
parameters reflected injury; however, the BBB scores were
more variable and did not show consistent injury, as
compared to the SEP recordings. The BBB locomotor
scores divided the cohort into approximately two groups:
one in early stages of recovery and another in intermediate
stages of recovery despite the level of injury being iden-
tical. The statistical difference in SEP measures across
different time points was much stronger (p * 10-16) than
that for BBB measures (p * 10-6). Hence it is evident that
the SEP measures were more sensitive to the injury caused
due to focal inflammatory and demyelinating lesion in the
spinal cord than the BBB measures. Our results also sug-
gest that this may generally hold true in assessments of
therapeutic interventions designed to repair the injured
spinal cord.
As noted earlier, both the SEP amplitudes and the BBB
showed a marginal improvement on day 7 compared to day
4. This is related to acute phase of injury due to incision of
paravertebral muscles, laminectomy, and injury which
could amplify the effect of injury during the first few days.
These could subside in the subacute period and could result
into a marginal improvement in the outcome measures. As
noted, we have found these improvements from day 4 to
day 7 to be statistically insignificant in our experimental
cohort.
In conclusion, our study suggests that behavioral scoring
may not be sufficient to characterize injury in the dorsal
part of the spinal cord and hence, electrophysiological
measurements using SEP could be used as a complement to
the BBB outcomes. In addition, BBB may not estimate the
condition well during the very acute phase of injury or a
Fig. 4 Histological
comparisons between rat spinal
cord with EAE (top 3 panels)
and rat spinal cord with no
injury
Neurol Sci
123
few days after injury due to inflammation and surgery.
Although BBB is relatively easy to execute and takes a
short period of time, only 4 min, it also has an element of
subjectivity, whereas SEP is a more objective measurable
parameter with reliable technology. A weakness in EP
monitoring is that it requires additional electrode prepara-
tion, instrumentation, and signal processing. Nevertheless,
a major benefit is that the assessment methodology is easily
reproducible and can be easily standardized. Therefore we
assert that SEP should be integrated into studies where
injury and lesion is induced to the spinal cord and should
be used to supplement motor behavior assessment, when-
ever possible.
Acknowledgments The study was supported by the Maryland Stem
Cell Research Fund (Grants 2007-MSCRFII-0159-00 and 2009-
MSCRFII-0091-00) and the Johns Hopkins Project RESTORE fund
(Transverse Myelitis Research Project). The authors would also like
to thank Payal Patnaik, Sonny Dike and Michael Gorelik for pro-
viding services during the entire experiment.
References
1. National Multiple Sclerosis Society, http://www.nationalmssociety.
org/index.aspx
2. Poser CM, Paty DW, Scheinberg L, McDonald WI, Davis FA,
Ebers GC, Johnson KP, Sibley WA, Silberberg DH, Tourtellotte
WW (2004) New diagnostic criteria for multiple sclerosis:
guidelines for research protocols. Ann Neurol 13:227–231
3. Polman CH, Reingold SC, Edan G, Filippi M, Hartung HP,
Kappos L, Lublin FD, Metz LM, McFarland HF, O’Connor PW,
Wollheim MS, Thompson AJ, Weinshenker BG, Wolinsky JS
(2005) Diagnostic criteria for multiple sclerosis: 2005 revisions to
the ‘‘McDonald Criteria’’. Ann Neurol 58:840–846
4. Hart BA, Amor S, Jonker M (2004) Evaluating the validity of
animal models for research into therapies for immune-based
disorders. Drug Discov Today 9:517–524
5. Kerschensteiner M, Stadelmann C, Buddeberg BS, Merkler D,
Bareyre FM, Anthony DC, Linington C, Bruck W, Schwab ME
(2004) Targeting experimental autoimmune encephalomyelitis
lesions to a predetermined axonal tract system allows for refined
behavioral testing in an animal model of multiple sclerosis. Am J
Pathol 164:1455–1469
6. DeBoy CA, Zhang J, Dike S, Shats I, Jones M, Reich DS, Mori S,
Nguyen T, Rothstein B, Miller RH, Griffin JT, Kerr DA,
Calabresi PA (2007) High resolution diffusion tensor imaging of
axonal damage in focal inflammatory and demyelinating lesions
in rat spinal cord. Brain 130:2199–2210
7. Basso D, Beattie M, Bresnahan J (1995) A sensitive and reliable
locomotor rating scale for open field testing in rats. J Neuro-
trauma 12:1–21
8. Basso D, Beattie M, Bresnahan J (1996) Graded histological and
locomotor outcomes after spinal cord contusion using the NYU
weight-drop device versus transaction. Exp Neurol 139:244–256
9. Young W (1996) Secondary injury mechanisms in acute spinal
cord injury. J Emerg Med 11:13–22
10. Finnerup NB, Gyldensted C, Fuglsang-Frederiksen A, Bach FW,
Jensen TS (2004) Sensory perception in complete spinal cord
injury. Acta Neurol Scand 109:194–199
11. Agrawal G, Thakor NV, All AH (2009) Evoked potential versus
behavior to detect minor insult to spinal cord in rat model. J Clin
Neurosci 16(8):1052–1055
12. Dawson GD (1947) Cerebral response to electrical stimulation of
peripheral nerve in man. J Neurol Neurosurg Psychiatry 10:137–
140
13. Grundy BL (1984) Intraoperative monitoring of sensory evoked
potentials. In: Nodar RH, Barber C (eds) Evoked Potentials II.
Butterworth, Boston, p 624
14. Chiappa K (1990) Evoked potentials in clinical medicine, 2nd
edn. Raven Press, New York
15. Nuwer MR (1998) Spinal cord monitoring with somatosensory
techniques. J Clin Neurophysiol 15:183–193
16. Nuwer MR (1998) Fundamentals of evoked potentials and com-
mon clinical applications today. Electroencephalogr Clin Neu-
rophysiol 106:142–148
17. Sherman DL, Atit MK, Geocadin RG, Venkatesha S, Hanley DF,
Natarajan AL, Thakor NV (2002) Diagnostic instrumentation for
neurological injury. IEEE Instru Measur Mag 28–35
18. Brown RH, Nash CL, Berilla JA, Amaddio MD (1984) Cortical
evoked potential monitoring: a system for intraoperative moni-
toring of spinal cord function. Spine 9:256–261
19. York DH, Chabot RJ, Gaines RW (1987) Response variability of
somatosensory evoked potentials during scoliosis surgery. Spine
12:864–876
Neurol Sci
123