Portable Custom Built Device for Thermal Sensitivity Assessment

An Auxiliary Tool to Characterize the Neuropathic Pain following Spinal Cord

Injury

Renato Varoto

, Fábio Casagrande Hirono

, Fernando Ometto Zorzenoni

Ricardo Yoshio Zanetti Kido

, William Barcellos

and Alberto Cliquet Jr.

1,2

Department of Electrical Engineering, University of São Paulo (USP), São Carlos, Brazil

Department of Orthopedics and Traumatology, University of Campinas (UNICAMP), Campinas, Brazil

Keywords: Thermal Sensitivity Assessment, Neuropathic Pain, Spinal Cord Injury.

Abstract: Neuropathic pain is characterized to arise without stimulation of nociceptors, but due to injury or

dysfunction of Peripheral and Central Nervous Systems. It involves altered mechanisms of impulse

transmission in somatosensory pathways, causing abnormal sensations. Quantitative sensory testing, by the

detection of thermal stimuli, is a method used to characterize and study the neuropathic pain. Therefore, this

work describes the development and application of portable custom built device for cutaneous thermal

sensitivity assessment in spinal cord injured subjects (SCIS). Using method of levels, the assessment was

applied in healthy subjects and SCIS with and without neuropathic pain. The thresholds determined for

healthy subjects during thermal sensitivity assessment are consistent and other results provided by clinical

trials are according to previous works, demonstrating the device feasibility as an auxiliary tool for

neuropathic pain study.

1 INTRODUCTION

Spinal cord injury (SCI) causes disruption of nerve

fibres that transmit ascending sensory and

descending motor information. This disruption

causes losses in the transmission of sensory-motor

information across the site of the lesion, resulting in

considerable physical and emotional consequences

for individual (Maynard Jr. et al., 1997); (Eng and

Miller, 2006). Sensory-motor dysfunctions occur in

the parts of the body innervated by areas below the

site of the lesion, being characterized by paralysis,

altered sensation and weakness (Raineteau and

Schwab, 2001).

Spinal cord injured subjects (SCIS) also suffer

other disorders and numerous secondary pathologies

such as losses of bowel and bladder functions,

pressure ulcers, spasticity, gastrointestinal and

sexual dysfunctions and heterotopic ossification

(Kaplan et al., 1991); (Eng and Miller, 2006);

(Verschueren et al., 2011). However, one of the

major problems following SCI is the neuropathic

pain (Bonica, 1991).

Neuropathic pain is characterized to arise

without stimulation of nociceptors (sensory pain

fibres that detect tissue damage by physical,

chemical or thermal phenomena), but due to injury

or dysfunction of Peripheral and Central Nervous

Systems. Thus, neuropathic pain is an aggravating

for the already weakened patient, imposing severe

limitations in performing the activities of daily

living (Richards et al., 1980); (Summers et al.,

1991).

The pathophysiology of neuropathic pain

involves altered mechanisms of impulse

transmission in somatosensory pathways, so that

axonal injury leads to a gain in excitatory

transmission, in other words, there is a massive

axonal input. It results from an axonal

hyperexcitability, with the generation of ectopic

electrical impulses, causing abnormal sensations

(Catafau and Bosque, 2003).

In SCIS, partially preserved pathways

spinothalamic tract may be the local generator of

pain (Wasner et al., 2008). Fibres Aδ and C present

little myelin and follow the column via anterolateral

spinothalamic tract. These fibres are the main

components of the fibres that lead thermal sensitivity

(Kirillova et al., 2011) Thus, the thermal sensitivity

Varoto R., Casagrande Hirono F., Ometto Zorzenoni F., Yoshio Zanetti Kido R., Barcellos W. and Cliquet Jr. A..

Portable Custom Built Device for Thermal Sensitivity Assessment - An Auxiliary Tool to Characterize the Neuropathic Pain following Spinal Cord Injury.

DOI: 10.5220/0004225900280034

In Proceedings of the International Conference on Biomedical Electronics and Devices (BIODEVICES-2013), pages 28-34

ISBN: 978-989-8565-34-1

 2013 SCITEPRESS (Science and Technology Publications, Lda.)

follows the same neurological path of the pain.

Some methods are applied to characterize and

study the neuropathic pain, for example, McGill

Pain Questionnaire, quantitative sensory testing

(QST) and somatosensory evoked potential

(Finnerup et al., 2003).

The McGill Pain Questionnaire is an instrument

that evaluates qualitatively and quantitatively pain,

providing quantitative measures of clinical pain that

can be treated statistically (Melzack, 1975). Pimenta

and Teixeira (1996) adapted (translation and

validation) the questionnaire to Portuguese. The

present pain intensity (PPI) is the number chosen by

the subject at the time of administration of the

questionnaire, ranging from 0 (no pain) to 5

(excruciating). The pain rating index based on the

subjects’ mean scale values (PRI(S)) obtained by

Melzack and Torgerson (1971) is described as the

sum of all values of words chosen by subject for all

categories (sensory, affective, evaluative, motor and

miscellaneous). And other important value is the

number of words chosen (NWC) that is the sum of

all words chosen by the subject (Melzack, 1975).

QST assess and quantify sensory function in

subjects with losses in the neurological system,

measuring the detection threshold of tactile,

vibratory, thermal or painful stimuli (Shy et al.,

2003); (Finnerup et al., 2003). Especially for thermal

stimuli, some equipments utilize the Peltier effect, in

which the intensity and direction of electrical current

controls the surface temperature of a test electrode.

The skin was touched by the electrode and the

subject reports the sensation in relation to the

temperature (Shy et al., 2003); (Kenshalo and

Bergen, 1975); (Finnerup et al. 2003).

This paper describes the development and

application of portable custom built device for

cutaneous thermal sensitivity assessment based on

Peltier effect. This device is designed for quick and

practice assessment of thermal sensitivity in SCIS,

representing an auxiliary tool for neuropathic pain

study. Using method of levels, the device was used

in healthy subjects and SCIS with and without

neuropathic pain. In general, the obtained results

were compared with previous works to verify the

device feasibility.

2 MATERIALS AND METHODS

About this work, instrumentation development was

done at Laboratory of Biocybernetics and

Rehabilitation Engineering - USP, and clinical

application was performed at Laboratory of

Biomechanics and Rehabilitation of the Locomotor

System – UNICAMP.

Basically, portable custom built device is

composed by microcontroller, thermoelectric

module and temperature transducer. The

microcontroller associated with amplifier circuits

offers electrical energy to supply the thermoelectric

module and provides information about device

operation condition. Furthermore, it allows setting

the probe operating temperature (Figure 1).

Figure 1: Block diagram of the device.

2.1 Thermoelectric Module,

Temperature Transducer and

Probe Assembly

The thermoelectric module used was a solid state

heat pump (Melcor Corporation, Trenton, NJ, USA),

based on Peltier effect. This heat pump contains 66

thermocouples, being able to transfer until 3.56W of

heat from cold to hot faces (Q

max

); it results in the

maximum temperature difference of the 67

between two faces (∆T

max

) with low power

consumption - input electrical current (I

max

) of 0.8A

and dc voltage (V

max

) of 7.98V – to achieve ∆T

max

Temperature transducer with analog output based

on semiconductor junctions was used to monitor the

temperature of thermoelectric module. The

transducer used was LM35 (National Semiconductor

Corporation, Santa Clara, CA, USA) that is a

precision integrated-circuit temperature transducer,

whose output voltage is linearly proportional to the

Celsius temperature. The LM35 does not require

external calibration to provide readouts with

accuracies of ±0.75

C over a range -55 to +150

Other important features of LM35 make it suitable

for control circuits as low output impedance, very

low self-heating and sensitivity of 0.01V/

The probe is composed by aluminium touch plate

(16x16mm), thermoelectric module, LM35

transducer, heat sink and auxiliary fan. In the first

stage of probe assembly, LM35 was coupled to the

touch plate by aluminium clamp, and the

PortableCustomBuiltDeviceforThermalSensitivityAssessment-AnAuxiliaryTooltoCharacterizetheNeuropathic

PainfollowingSpinalCordInjury

thermoelectric module was fixed on heat sink

(Figure 2). Touch plate with LM35 was fixed on

thermoelectric module by thermal compound, in the

second stage. Moreover, the assembly was placed in

a suitable case.

Figure 2: First stage of probe assembly.

2.2 Electronic Apparatus

The microcontroller used was PIC18F252

(Microchip Technology Inc., Chandler, AZ, USA)

and it was programmed to use pulse-width

modulation (PWM) as technique for controlling

power to thermoelectric module. It compares the

desired temperature of the probe with the actual

temperature and, in accordance with this difference,

adjusts the duty cycle of PWM signal.

An active low-pass filter was used to generate a

dc signal from 0 – 5V PWM signal. This filter was

configured in Sallen-Key topology of second order

with cut-off frequency of approximately 35 Hz and

quality factor of 0.5.

During bench tests with the probe, dc voltages of

thermoelectric module to ensure that the temperature

probe reached 0 and 60

C were determined. The

voltage polarity was defined so that a negative

voltage decreases the probe temperature and a

positive one would increase the temperature. These

voltages were -7.2V and 4.8V, respectively.

Therefore, it was necessary to convert 0 – 5V dc

signal (Vin) obtained from PWM into asymmetric

bipolar dc signal (-7.2V – 4.8V) (Vout) required by

thermoelectric module. This conversion was done by

differential amplifier using an operational amplifier

as active component, whose inverting input was held

at 5V. Thus, an amplifier with linear transfer

characteristic was developed (1), resulting in proper

range of signal.

Vout=2.435*Vin - 7.353 (1)

Although the range of signal was appropriate, the

differential amplifier was not able to provide the

required electrical power (up to 5.76W). Thus, a

push-pull stage with MOSFETs was applied. These

components exhibit a negative thermal coefficient,

in other words, its electrical conductivity decreases

with increasing temperature, protecting the

electronic circuit.

The output of push-pull stage with unitary gain

was also used as feedback signal to differential

amplifier. This strategy reduces signal distortion,

providing a linear signal to thermoelectric module.

Whole electronic circuit including the thermoelectric

module is powered by two batteries (12V, 5Ah).

2.3 Clinical Trials

Twenty SCIS were recruited to participate in this

work, and they were divided into two groups, pain

SCIS (P) and non-pain SCIS (NP). Subjects were

classified according to the American Spinal Cord

Injury Association (ASIA) Impairment Scale (AIS)

(Maynard Jr. et al., 1997); (Dahlberg et al., 2005);

(Wolfe and Hsieh, 2006); (Kirshblum et al., 2011).

Control group (CT) was formed by 10 healthy

subjects (Table 1).

Table 1: Subjects characteristics.

P NP CT

Age (year)*

39.3(12.1) 35.4(12.9) 27.2(10.6)

Gender

(Male/Female)

9/1 7/3 9/1

Body mass

(kg)*

69.2(10.1) 71.5(14.7) 77.6(13.6)

Height (m)*

1.74(0.07) 1.72(0.09) 177.3(4.8)

Neurological

lesion level

(Cervical/

Thoracic)

7/3 3/7 -

AIS (A/B/C/D)

8/2/0/0 7/2/1/0 -

*Values in mean(SD)

Inclusion criteria for P group were lesion level

above T12 with central neuropathic pain after

traumatic SCI. Exclusion criteria were based on the

presence of any other pain different from central

neuropathic pain such as nociceptive or peripheral

neuropathic pain; or subjects that were under

analgesic treatment.

In relation to NP group, inclusion criteria were

lesion level above T12 without central neuropathic

pain and spontaneous dysesthesia.

To study the sensitivity of pain and losses of

sensory pathways following SCI, McGill Pain

Questionnaire (Portuguese version) and thermal

sensitivity assessment were applied.

For McGill Pain Questionnaire, the values used

BIODEVICES2013-InternationalConferenceonBiomedicalElectronicsandDevices

for data analysis were PRI, NWC and PPI.

The thermal stimuli were applied to the dominant

leg at a point 100mm distal from the patella, in the

anterolateral side of the leg, corresponding to the L5

dermatome. For temperature range from 30

C to

C, with increment of 5

C, the skin was stimulated

by probe (aluminium touch plate, specifically) by

over 3s. Subsequently, the temperature range was

from 30

C to 0

C, with decrement of 5

C. Warm and

cold thresholds (temperature at which the patient

feels the stimulus) and pain thresholds were

recorded using the method of levels (Shy et al.,

2003). For each subject, three measurements with

interstimulus interval of 3 – 6s were used to

calculate the thresholds.

3 RESULTS

Figure 3 shows portable custom built device for

thermal sensitivity assessment; probe and control

module.

Figure 3: Portable custom built device for thermal

sensitivity assessment.

On the front panel, the control module has two

pushbuttons that set the desired probe temperature;

the red pushbutton (+) increases probe temperature

of 1

C while the black one (-) decreases it of 1

C, at

range of 0

C to 70

C. This desired probe

temperature and the instantaneous one are shown on

the smaller green display and larger red display,

respectively. When temperatures become equal,

LED turns on, indicating that probe is ready to use.

Besides, the front panel has a toggle switch for the

auxiliary fan and a DB9 connector for the probe

cable.

According to the McGill Pain Questionnaire

applied to the P group, half of subjects feel pain at

injury level and half of them below the injury level.

Figure 4 shows the relation between reported words

and number of subjects for each group of

questionnaire; and table 2 presents the scores for

each variable.

Figure 4: Relation between reported words and number of

subjects for each group of McGill Pain Questionnaire

(Portuguese version).

Table 2: McGill pain questionnaire scores.

Variables McGill Pain Questionnaire scores

PRI

28.5(13.7)

NWC

12.7(3.0)

PPI

3.2(1.4)

*Values in mean(SD)

During thermal sensitivity assessment, three

subjects detected cold, two detected warm and one

detected pain due to heat, in the NP group (Table 3).

Furthermore, none reported pain due to cold and one

subject (AIS C) presented muscle spasms with

stimulus of 50

Table 3: Thermal sensitivity in the NP group.

Subjects (AIS)

Detected threshold ▲

Cold Pain due to cold Warm Pain due to heat

1(A) - - - -

2(B) - - - -

3(B) ▲ - ▲ ▲

4(A) - - - -

5(A) ▲ - - -

6(A) - - - -

7(A) - - - -

8(A) - - - -

9(A) ▲ - ▲ -

10(A) - - - -

In the P group, four subjects detected cold, one

detected pain by cold at 5°C and seven subjects

detected warm. Four subjects detected pain due to

heat and heat pain tolerance, with three subjects (all

PortableCustomBuiltDeviceforThermalSensitivityAssessment-AnAuxiliaryTooltoCharacterizetheNeuropathic

PainfollowingSpinalCordInjury

AIS A) presenting muscular spasms at 55°C (Table

4). One subject (AIS B) felt dysesthesia at the

stimulation site with stimuli of 0ºC and 45°C.

Another subject (AIS A) detected warm all around

the knee (not only at the stimulation site) detecting

warm at 0°C. And one subject (AIS A) felt a non

specific vibration in the L5 dermatome with 45°C in

the right leg.

Table 4: Thermal sensitivity in the P group.

Subjects (AIS)

Detected threshold ▲

Cold Pain due to cold Warm Pain due to heat

1(A) - - ▲ ▲

2(B) - - ▲ -

3(B) ▲ - ▲ ▲

4(A) - - - -

5(A) - - ▲ -

6(A) ▲ - ▲ ▲

7(A) ▲ - ▲ -

8(A) - - - -

9(A) - - - -

10(A) ▲ ▲ ▲ ▲

Table 5 indicates cold (C), warm (W), pain due

to heat (HP) and heat pain tolerance (HPT)

thresholds for each group.

Table 5: Temperature thresholds for each group.

Threshold(

Group

P NP CT

15(9.8) 20.6(9.5) 19.2(5.2)

38.3(11.9) 38.3(2.6) 36.4(3.3)

47.9(5.4) 50(0) 50.8(2.6)

HPT

48.3(4.9) 50(0) 52.8(3.1)

Values in mean(SD)

4 DISCUSSION

For touch plate construction, aluminium, copper and

stainless steel were available. The choice was based

on coefficient of thermal conductivity and oxidation

resistance of metals. According to the coefficient of

thermal conductivity, copper (398W/mK) is a better

conductor than aluminium (247W/mK) and stainless

steel (15.9W/mK), but stainless steel presents high

oxidation resistance. Therefore, aluminium was

chosen because presents high coefficient of thermal

conductivity and intermediate oxidation resistance

(Callister Jr., 2001).

These properties associated to the low mass and

small dimension of touch plate (0.7g) enabled a fast

thermal equilibrium between both plate surfaces.

Thus, the surface temperature acquired by the

transducer is the same of surface dedicated to the

touch.

Reach and stabilization of desired probe

temperature can be attributed to the technique for

controlling power to thermoelectric module and the

use of heat sink and auxiliary fan. For each

increment or decrement of 5

C, this strategy allowed

temperature stabilization in around 5s during clinical

trials.

In relation to the technique for controlling power

to thermoelectric module, another alternative based

on the use of PWM signal and an H-bridge can be

applied, replacing low-pass filter and differential

amplifier. However, this configuration provides a dc

voltage range from -12V to 12V for the PWM duty

cycle of 0 and 100%, respectively; values which are

not in accordance with the asymmetric bipolar dc

signal (-7.2V – 4.8V) required by the thermoelectric

module to operate in proper temperature range (0 –

C). Thus, the PWM duty cycle should be limited

between values of 20% and 70%, also avoiding

damage to the module since this operates at a

maximum dc voltage of 7.98V. Therefore, the use of

the low-pass filter and the differential amplifier is

more appropriate to the objectives of this work.

Generally, in healthy subjects, the activity of

cold-sensitive neurons increases below 35

C, and

maximum cutaneous cold sensitivity is around 25

while cold fibre activity is ceased at temperatures

below 12

C. The firing rates of warm-sensitive

neurons increase above 25

C, and their range of

thermosensitivity extends from 35

C to 43

Temperatures above 43

C and below 12

C cause

pain, whose stimulus is transmitted by Aδ and C

fibres. Furthermore, nociceptive heat activates Aδ

fibres around 43

C, while temperatures above 52

activate C fibers. Cold stimuli below 12

C also

cause pain and, in addition, nociceptives heat and

cold are transmitted by polymodal C fibers (Nomoto

et al., 2004; Schepers and Ringkamp, 2010).

Therefore, the thresholds determined for CT group

during thermal sensitivity assessment using portable

custom built device are consistent (Table 3).

In relation to SCIS, P group was more sensitive

to thermal stimuli than NP group, where 70% of

subjects in P group detected some kind of thermal

sensitivity against 30% of NP group (Tables 3 and

4). This difference between P and NP groups is in

agreement with the study of Wasner et al. (2008). In

this study, it was reported that there is some

preservation of the spinothalamic tract in pain SCIS

greater than in non pain SCIS, which may be

involved in the development of neuropathic pain.

BIODEVICES2013-InternationalConferenceonBiomedicalElectronicsandDevices

From the three subjects who experienced thermal

stimuli in NP group, only one presents complete

injury. For the P group, two subjects are AIS B and

five are AIS A.

This finding can be justified through theories that

explain why subjects with complete SCI have some

sensibility, characterizing the discomplete injury.

Dimitrijevic (1988) and Sherwood, Dimitrijevic and

Mckay (1992) found motor remnants in complete

SCIS due to a neural control. Thus, discomplete

injury is an incomplete injury that fits the AIS

criteria for grade A. Moreover, some subjects with

complete SCI can present some semblance of

sensibility, which can be evoked below the level of

injury due to incomplete injuries in the

spinothalamic tract. These subjects have subclinical

functions of ascending and descending tracts

(Finnerup et al., 2004).

5 CONCLUSIONS

Due to feedback control, the custom built portable

device provides easy temperature control with

resolution of 1

C. The device is simple to build and

can stabilize its temperature in about 5s for a 5

temperature change, therefore representing a simple

alternative for quick and practical assessment. It can

provide quantified information about sensory

performance of subjects, and the results obtained

from clinical trials are in accordance to previous

works, thus demonstrating the device feasibility for

thermal sensitivity assessment. Spinal cord injured

subjects that refer neuropathic pain are more

sensitive to thermal stimuli than patients do not

present neuropathic pain.

ACKNOWLEDGEMENTS

We thank the support by grants from São Paulo

Research Foundation (FAPESP) and National

Council for Scientific and Technological

Development (CNPq).

REFERENCES

Bonica, J. J., 1991. Introduction: semantic, epidemiologic,

and educational issues. In: Casey, K. L. Pain and

Central Nervous System Disease: the Central Pain

Syndromes. Raven Pr. New York: Raven Pr, pp 13–

29.

Callister Jr., W. D., 2001. Fundamentals of Material

Science and Engineering. 5

ed. New York: John

Wiley & Sons, Inc.

Catafau, S., Bosque, Q., 2003. Mecanismos

fisiopatológicos da dor neuropática, 1

ed. Madrid:

Medica Panamericana.

Dahlberg, A., Alaranta, H., Sintonen, H., 2005. Health-

related quality of life in persons with traumatic spinal

cord lesion in Helsinki. Journal of Rehabilitation

Medicine, 37, pp. 312–316.

Dimitrijevic, M. R., 1988. Residual motor functions in

spinal cord injury. Advances in Neurology, 47, pp.

138–155.

Eng, J. J., Miller, W. C., 2006. Rehabilitation: from

bedside to community following spinal cord injury

(SCI). In: Eng, J. J., Teasell, R. W., Miller, W. C.,

Wolfe, D. L., Townson, A. F., Aubut, J., Abramson,

C., Hsieh, J. T. C., Connolly, S. Spinal Cord Injury

Rehabilitation Evidence. Vancouver, pp. 16–29.

Finnerup, N. B., Johannesen, I. L., Fuglsang-Frederiksen,

A., Bach, F. W., Jensen, T. S., 2003. Sensory function

in spinal cord injury patients with and without central

pain. Brain, 126, pp. 57–70.

Finnerup, N. B., Gyldensted, C., Fuglsang-Frederiksen,

A., Bach, F. W., Jensen, T. S., 2004. Sensory

perception in complete spinal cord injury. Acta

Neurologica Scandinavica, 109, pp. 194–199.

Kaplan, S. A., Chancellor, M. B., Blaivas, J. G., 1991.

Bladder and sphincter behavior in patients with spinal

cord lesions. Journal of Urology, 146, pp. 113–117.

Kenshalo, D. R., Bergen, D. C., 1975. A device to

measure cutaneous temperature sensitivity in humans

and subhuman species. Journal of Applied Physiology,

39, pp. 1038–1040.

Kirillova, I., Rausch, V. H., Baron, R., Jänig, W., 2011.

Mechano- and thermosensitivity of injured muscle

afferents. Journal of Neurophysiology, 105, pp. 2058–

2073.

Kirshblum, S. C., Burns, S. P., Biering-Sorensen, F.,

Donovan, W., Graves, D. E., Jha, A., Johansen, M.,

Jones, L., Krassioukov, A., Mulcahey, M. J., Schmidt-

Read, M., Waring, W. 2011. International standards

for neurological classification of spinal cord injury

(Revised 2011). The Journal of Spinal Cord Medicine,

34, pp. 535–546.

Maynard Jr., F. M., Bracken, M. B., Creasey, G., Ditunno

Jr, J. F., Donovan, W. H., Ducker, T. B., Garber, S. L.,

Marino, R. J., Stover, S. L., Tator, C. H., Waters, R.

L., Wilberger, J. E., Young, W., 1997. International

standards for neurological and functional classification

of spinal cord injury. Spinal Cord, 35, pp. 266–274.

Melzack, R., 1975. The McGill Pain Questionnaire: major

properties and scoring methods. Pain, 1, pp. 277–299.

Melzack, R., Torgerson W. S., 1971. On the language of

pain. Anesthesiology, 422, pp. 50–59.

Nomoto, S., Shibata, M., Iriki, M., Riedel, W., 2004. Role

of afferent pathways of heat and cold in body

temperature regulation. International Journal of

Biometeorology, 49, pp. 67–85.

Pimenta, C. A. de M., Teixeira, M. J., 1996. Questionário

PortableCustomBuiltDeviceforThermalSensitivityAssessment-AnAuxiliaryTooltoCharacterizetheNeuropathic

PainfollowingSpinalCordInjury

de dor McGill: proposta de adaptação para a língua

portuguesa. Revista da Escola de Enfermagem da

USP, 30, pp. 473–483.

Raineteau, O., Schwab, M. E., 2001. Plasticity of motor

systems after incomplete spinal cord injury. Nature

reviews. Neuroscience, 2, pp. 263–273.

Richards, J. S., Meredith, R. L., Nepomuceno, C., Fine, P.

R., Bennett, G., 1980. Psychosocial aspects of chronic

pain in spinal cord injury. Pain, 8, pp. 355–408.

Schepers, R. J., Ringkamp, M., 2010. Thermoreceptors

and thermosensitive afferents. Neuroscience and

Biobehavioral Reviews, 34, pp. 177–184.

Sherwood, A. M., Dimitrijevic, M. R., Mckay, W. B.,

1992. Evidence of subclinical brain influence in

clinically complete spinal cord injury: discomplete

SCI. Journal of the Neurological Sciences, 110, pp.

90–98.

Shy, M. E, Frohman, E. M., So, Y. T., Arezzo, J. C.,

Cornblath, D. R., Giuliani, M. J., Kincaid, J. C.,

Ochoa, J. L., Parry, G. J., Weimer, L. H., 2003.

Quantitative sensory testing. Neurology, 60, pp. 898–

904.

Summers, J. D., Rapoff, M. A., Varghese, G., Porter, K.,

Palmer, R. E., 1991. Psychosocial factors in chronic

spinal cord injury pain. Pain, 47, pp. 183–189.

Verschueren, J. H. M., Post, M. W. M., de Groot, S., van

der Woude, L. H. V., van Asbeck, F. W. A., Rol, M.,

2011. Occurrence and predictors of pressure ulcers

during primary in-patient spinal cord injury

rehabilitation. Spinal Cord, 49, pp. 106–112.

Wasner, G., Lee, B. B., Engel, S., Mclachlan, E., 2008.

Residual spinothalamic tract pathways predict

development of central pain after spinal cord injury.

Brain, 131, pp. 2387–2400.

Wolfe, D. L., Hsieh, J. T. C., 2006. Rehabilitation practice

and associated outcomes following spinal cord injury.

In: Eng, J. J., Teasell, R. W., Miller, W. C., Wolfe, D.

L., Townson, A. F., Aubut, J., Abramson, C., Hsieh, J.

T. C., Connolly, S. Spinal Cord Injury Rehabilitation

Evidence. Vancouver, pp. 44–90.

BIODEVICES2013-InternationalConferenceonBiomedicalElectronicsandDevices