An electromyographic analysis of the deep cervical flexor muscles in performance of craniocervical flexion.

The cervical spine is surrounded by a complex arrangement of muscles that contribute to control of the head and neck. We believe the most important part of this system is the deep vertebral muscles--the longus colli and longus capitis muscles. The morphological design of these muscles aids their ability to provide support of the cervical lordosis and the cervical joints. (1-4) Using a computer model, Winters and Peles (4) showed that regions of local segmental instability were produced if only the large, more superficial muscles of the neck were simulated to produce movement, particularly in nearly upright or neutral positions. Muscle activity of the deep muscles was required in conjunction with activity of the larger muscles to stiffen or stabilize the segments, especially in functional mid-ranges, which are common working postures. We believe this knowledge provides support for a focus on these muscles in the rehabilitation process.

The deep cervical flexor (DCF) muscles have been studied using techniques such as computer modeling (3, 4) histological analyses, (5) and imaging studies. (1) There have been few attempts, however, at obtaining a direct measure of DCF muscle activity via electromyography (EMG). Inaccessibility of the DCF muscles has been the main limitation, preventing direct measurement. Surface electrodes can be used to measure the superficial cervical muscles; however, their use is restricted when the muscle of interest is deeply positioned, such as the longus colli and longus capitis muscles of the cervical spine. Several decades ago, indwelling, fine-wire electrodes were used to measure these deep muscles. (6, 7) Both EMG studies were limited to subjects without known impairment or pathology, and, as a result, they provided very little information that could be used in the development of interventions for patients with neck pain.

Given the complexity and proximity of nearby structures such as the trachea, carotid artery, vagus nerve, and lymphatics, the technique of fine-wire EMG is extremely difficult and highly invasive and thus unsuitable for wider application.

We developed an approach that allows a direct measure of the DCF muscles to be obtained. The apparatus used in this approach, which was developed by one of the authors (PD), consists of electrode contacts attached to a suction catheter that are placed on the posterior oropharyngeal wall using a nasopharyngeal application. The DCF muscles he directly posterior to the oropharyngeal wall (Fig. 1). This proximity provides a location to make recordings of the DCF muscles via the mucosal wall, without requiring intramuscular recording techniques.

[FIGURE 1 OMITTED]

The purpose of our study was to evaluate this EMG technique for the examination of the DCF muscles. The craniocervical flexion (CCF) test (8, 9) was selected to evaluate the EMG signals. This test involves performing 5 stages of increasing CCF range of motion (ROM) (head nod action), which is the anatomical action of the DCF muscles. In our opinion, a measure of the DCF muscles during performance of tasks involving increasing CCF ROM should demonstrate increasing EMG amplitude, although it may not necessarily be linear, from these muscles with increasing effort. The reliability of EMG measurements obtained from the DCF muscles also was assessed. A reliable direct measurement of the DCF muscles could provide the opportunity for further research into the role of impairments in these muscles, which is believed to exist in people with neck pain. (8-10)

Method

Subjects

After giving informed consent, 10 volunteers (4 men, 6 women) between 21 and 53 years of age (X = 30.7, SD = 10.3) participated in this study. Subjects were included if they were free of neck pain at the time of testing, had no past history of orthopedic disorders affecting the neck, and had no history of neurological disorders. Each subject was screened according to exclusion criteria that were based on the contraindications and precautions for the use of Xylocaine spray local anesthetic(11, *) and for the use of the nasopharyngeal suctioning technique. (12) The spray and the suctioning technique are required for the surface EMG measure of the deep cervical flexor muscles.

Instrumentation and Measurements

Myoelectric signals were detected from the DCF muscles on the left side using bipolar electrodes (Fig. 2). The apparatus consisted of custom-made, bipolar, silver wire electrode contacts (dimensions: 2 x 0.6 mm, inter-electrode distance: l0 mm) that were attached to a suction catheter (size 10FG) with a heat-sealed distal end. This apparatus was inserted through the nose to the posterior oropharyngeal wall. Although activity of adjacent muscles may contribute to the signal, we believe this activity was minimized by fixing the electrode to the mucosa with suction pressure of 30 mm Hg through a portal between the 2 electrode contacts and by placing the electrode at the level of the uvula (approximately the level of the C2-3 intervertebral disk), which is the level at which the longus colli muscle has its greatest cross-sectional area. (13) Each electrode-tipped catheter was individually packed and sterilized using standard gas sterilization procedures.

[FIGURE 2 OMITTED]

Recordings of EMG activity were made bilaterally from the sternal head of the sternocleidomastoid (SCM) and anterior scalene (AS) muscles using disposable bipolar surface electrodes. [dagger] The electrodes were positioned at one third of the distance from the sternal notch to the mastoid process at the distal end overlying the muscle belly (14) following careful skin preparation. (15) A strap located around the wrist provided a ground reference. Signals were passed through a 20- to 450-Hz bandwidth filter (40 dB/decade slope on each side), sampled at 2,048 Hz, and amplified (ASE16 16-channel amplifier [double dagger]). The samples were digitized by a 12-bit, analog-to-digital converter (DAQ PCI 6024E-[10.sup.[section]]) and stored on a disk of a personal computer.

To obtain a measurement of EMG signal amplitude, the 1-second maximum root mean square (1sRMS) was calculated for each EMG trace using a custom-designed software program. [parallel] A reference voluntary contraction of CCF and cervical flexion should have reflected the combined actions of the deep and superficial cervical flexor muscles. This was done in the supine position in order to normalize the data. The 1sRMS values obtained during each stage of the CCF test for both the deep and superficial cervical flexor muscles were normalized by expressing them as a percentage of the maximum 1sRMS values obtained during the reference voluntary contraction.

The CCF test consists of 5 incremental movements of increasing CCF ROM. (16) Performance is guided by visual feedback from an air-filled pressure sensor (#) positioned suboccipitally. During the CCF test, subjects were required to perform gentle nodding motions of CCF that progressed in range to increase the pressure by 5 incremental levels, with each increment representing 2 mm Hg. The starting pressure was 22 mm Hg, and the ending pressure was 30 mm Hg. Recording of pressure measurements was made by connecting the pressure bag and pump to a pressure transducer. Electrical signals from the pressure transducer were amplified and relayed to a visual feedback device and to the integrated amplifier, the analog-to-digital converter, and the data storage system. ** The visual feedback device consisted of a voltmeter, which was marked in 2-mm Hg increments from 20 mm Hg to 30 mm Hg and which was calibrated to display the pressure in the pressure bag, based on the pressure transducer output. Sampling frequency for pressure measures was 1,000 Hz. The pressure traces were reviewed by the chief investigator at the completion of each trial to ensure that the subjects reached each pressure target and maintained the pressure steady on the target for the duration of recording.

The range of CCF that subjects obtained for each stage of the test was measured from a lateral photograph taken with a digital camera (PowerShot 100 DIGITAL IXUS [double dagger]) and using custom-designed analytical software for angle measurements (LabVIEW 6.0[i.sup.[section]). This method has a high level of reliability. (16)

Experimental Procedure

Subjects were positioned on a plinth in a supine crook-lying position. The starting position was standardized by placing the craniocervical and cervical spines in a position in which the subjects' forehead and chin were horizontal and in an imaginary line that was parallel to the plinth and extended from the tragus of the ear and bisected the neck longitudinally. (16)

The pressure biofeedback unit was placed suboccipitally behind the subjects' neck and set to a baseline pressure of 20 mm Hg. Subjects were instructed in how to do CCF, and they practiced the head-nodding action to progressively target (reach the incremental targets) and hold the 5 pressure levels for 10 seconds between 22 mm Hg and 30 mm Hg. Any substitutions such as neck retraction were identified by the chief investigator using visual inspection of subject performance from the lateral side and were discouraged. The combined movement of CCF and cervical flexion (head lift with chin rock) also was practiced in preparation for the reference voluntary contraction.

The digital camera was positioned on a tripod parallel to the subjects' head and neck region at a distance of 80 cm. Anatomical markers were positioned on the tragus of the ear, the mental protuberance of the mandible, and the lateral aspect of the neck--7 cm inferior to the mastoid process (16) (Fig. 3). Markers were fixed with double-sided medical tape. An initial photograph was taken of each subject in the starting position, followed by a photograph taken after the subject achieved full range of head nod (CCF) while in the supine position.

[FIGURE 3 OMITTED]

Surface electrodes were positioned bilaterally over the sternal head of the SCM and AS muscles. For insertion of the nasopharyngeal electrode, 3 metered doses of Xylocaine spray were administered to the nasal mucosa via the nostril by positioning the disposable nozzle inside the external aperture of the nostril and orienting the tip slightly medially toward the nasal septum. Subjects were asked to hold their breath during the application, to avoid breathing out against the spray. An additional 3 metered doses of anesthetic were administered through the mouth to the posterior oropharyngeal wall on the same side. A tongue depressor was used by the person administering the spray in order to provide greater visibility of the throat. Following a standard nasopharyngeal suction application, (17) the catheter was introduced. Subjects were asked to open their mouth as the tip of the electrode approached the oropharynx. This allowed the investigator to observe the position of the electrode contacts so that correct orientation and placement of the electrodes could be achieved. The electrodes were positioned directly onto the posterior oropharyngeal wall at the level of the uvula. This placement of the electrodes was designed to ensure correct orientation along the fibers of the DCF muscles, approximately 1 cm lateral to the midline. Once this position was achieved, suction was applied to maintain electrode-mucosa contact.

For the normalization procedure, subjects were instructed to perform the combined movement of CCF and cervical flexion. This movement consisted of a full CCF chin nodding action, followed by cervical flexion to lift the head so that it just cleared the plinth. In an effort to make sure there was correct movement, the investigator observed the motion. This contraction was maintained for 10 seconds and was repeated twice, with a rest period of 30 seconds between contractions. The highest value recorded over the 2 contractions became the reference 1sRMS value, allowing subsequent normalization. Subjects then performed the 5 stages of the CCF test, going from 22 mm Hg to 30 mm Hg and maintaining steady pressure on each target for 10 seconds. For each stage of the test, data collection commenced at the point when the subjects reached the pressure target. A digital image also was recorded at this point. An interval of 30 seconds was given between contractions, during which time an investigator checked each subject's head and neck position to ensure that the subject returned to his or her starting position. On completion of testing, the suction pressure was released and the catheter was removed gently.

Reliability

The reliability of normalized 1sRMS values obtained from the DCF muscles during the 5 stages of the CCF test was examined in 5 subjects (2 men, 3 women) between 22 and 53 years of age (X = 32.7, SD = 14.3). Each subject repeated the experimental protocol 3 times, 1 day apart over nonconsecutive days, performing the CCF test twice each day. Reliability estimates for the normalized 1sRMS values from the DCF muscles were obtained by evaluating intraclass correlation coefficients (ICCs) and the normalized standard error of the mean (SEM). Previous work examining the reliability of surface EMG measurements has identified the advantages of considering both the ICC and the normalized SEM. (18, 19) When n measurements are performed on each of m subjects, the n * m values have a mean ([micro]) and a variance ([[delta].sup.2]). The normalized SEM (100 * [delta]/[micro][square root of (n * m)]) gives information about the statistical range of values that [micro] would have in repeated experiments. A low normalized SEM indicates a reliable estimate of [micro] (eg, minimal experimental noise). Analysis of variance (ANOVA) provides information about the factors contributing to [[delta].sup.2] (days, trials, subjects, measurement conditions) regardless of the actual value of [[delta.sup.2]. Specifically, ICC is the portion of [[delta].sup.2] that is associated with differences among measured individuals and tells us if the variations of the measurements reflect subject-to-subject variations or random experimental noise, both of which are components of reliability. Throughout the text, the variances will be presented as a percentage of the total variance because we believe this increases the usefulness of the results.

Data Management and Statistical Analysis

A measure of the DCF muscles during increasing CCF ROM should demonstrate increasing, but riot necessarily linear, EMG amplitude from these muscles with increasing effort. A linear mixed model, fitting subjects as a random effect and with fixed stage effects, was therefore applied to determine whether a relationship existed between the amplitude of muscle activity and the 5 incremental stages of the CCF test. In addition, an analysis of contrasts was conducted using t tests to identify whether there were differences between normalized 1sRMS values for each muscle at each stage of the CCF test and whether there were differences in the EMG amplitude increase for each muscle between successive stages of the CCF test. A value of P<.05 was considered statistically significant. Preliminary analysis identified no difference between sides for the AS muscle (P = .5) and the SCM muscle (P = .8), allowing pooling of the data (N = 10).

The range of CCF was calculated by subtracting the angle at full head nod from that in the starting position. The ROM obtained at each stage of the CCF test was then expressed as a percentage relative to the full range of CCF obtained. Pearson correlation coefficients were calculated to examine the relationship between range of upper cervical flexion and normalized 1sRMS from each of the 3 muscles studied (N = 10).

Results

A positive linear relationship was evident between the normalized 1sRMS for DCF muscles and the incremental stages of the CCF test (F = 239.04, df = 36, P<.0001). For both the AS and SCM muscles, a linear relationship between normalized 1sRMS and test level also was identified (AS muscle: F = 100.32, df = 36, P<.0001; SCM muscle: F = 85.31, df = 36, P<.0001). Table 1 presents the mean and SEM for the normalized 1sRMS values for each muscle obtained across the 5 stages of the CCF test. Increases in normalized 1sRMS of the DCF muscles were identified among the 5 stages of the CCF test. Differences in the normalized 1sRMS of both the AS and SCM muscles over the successive stages of the test were evident between the stages of 24 mm Hg and 30 mm Hg only (Fig. 4).

[FIGURE 4 OMITTED]

The range of CCF increased with each progressive stage of the test (Tab. 1). Pearson correlation coefficients for the relationship between ROM and normalized 1sRMS for the DCF, SCM, and AS muscles were r = .56 (P<.001, confidence interval [CI] = .341-.729), .51 (P<.001, CI = .270-.690), and .45 (P<.001, CI = .202-.651), respectively, indicating that there was a positive relationship between the increasing angle of CCF and increasing EMG amplitude in the deep and superficial neck flexors.

Table 2 documents the results of ICC and normalized SEM for normalized 1sRMS values obtained from the DCF muscles. Low values of the within-subject normalized SEM were found for the normalized 1sRMS values for the DCF muscle (6.7%-10.3%), providing evidence of high reliability for these variables.

Discussion

The DCF muscles have been investigated using various indirect approaches (1,3,4); however, there have been few attempts to obtain a direct measure of DCF muscle activity because of the difficulty in measuring the activity of these muscles. In general, avoiding potential risks inherent in using fine-wire electrodes has limited the study of these muscles. The development of a nasopharyngeal surface electrode provided a method for obtaining direct measurements of the DCF muscles. The purpose of our study was to evaluate this technique for measuring the muscle activity of the DCF muscles by recording EMG signals during progressive stages of increasing CCF ROM during the CCF test. Despite the inclusion of small numbers of subjects because of the nature of testing, the results revealed a strong positive linear relationship between the normalized 1sRMSs for the DCF, SCM, and AS muscles and the incremental stages of the CCF test. Each stage of the test was accompanied by increasing EMG amplitude in the DCF muscles, which is consistent with the anatomical action of these muscles. Furthermore, increases in normalized 1sRMS of the DCF muscles were identified among the 5 stages of the CCF test. The superficial cervical flexors, the SCM and AS muscles, however, demonstrated an increase in EMG amplitude between the first 2 stages of the test (22 mm Hg and 24 mm Hg); no change in their magnitude of activity was identified. The results showed that the CCF test does not uniquely measure the activity of the DCF muscles, because activity was recorded in the SCM and AS muscles in each stage of the test. The results of this study also revealed a progressive increase in CCF ROM during each stage of the test. This increasing ROM correlates with increasing EMG amplitude for each of the 3 muscles examined, with the strongest correlation being identified for the DCF muscles.

Methodological Considerations

Recording of DCF muscle activity involved an EMG technique for the measurement of muscle activity of the DCF muscles. Further research, however, remains necessary to ensure whether cross talk from other nearby muscles, such as the superficial cervical flexor and submandibular muscles, affect recording. Given the location of the electrode and the small inter-electrode distance incorporated in the electrode design, we are confident that the majority of the signals detected are from the longus colli and longus capitis muscles. The low impedance associated with detection over a mucosal surface and fixation of the electrodes with suction, we believe, further assured quality of the myoelectric signals obtained. Other techniques have been used to record EMG activity through mucosa, such as recordings made from the crural fibers of the diaphragm through the esophageal wall. (20) We argue that our setup avoided the reported problems associated with movement of the electrode in that technique by fixation of the electrode with suction. The procedure was well tolerated by all subjects, and no side effects associated with the technique or anesthetic were reported.

The reference voluntary contraction of combined CCF and cervical flexion selected for normalization of EMG amplitude was not a true maximum voluntary contraction. As such, the amplitudes of DCF, SCM, and AS muscle activity cannot be accurately compared. Further studies, we believe, are necessary to investigate the contribution of the deep and superficial cervical flexor muscles during performance of the CCF test.

Reliability

The repeatability of normalized 1sRMS values for the DCF muscles obtained during the 5 stages of the CCF test demonstrate reliability or constancy. As previously documented, (18, 19, 21) when the between-subject variability is comparable or less than the within-subject variability, the degree of repeatability defined by the ICC becomes meaningless. The low values obtained for the between-and within-subject variability for the normalized 1sRMS values for the DCF muscles demonstrate that these variables were estimated with very high repeated-measure precision. However, these low values also show that there was very little variation in the normalized 1sRMS values obtained across subjects and trials, so the measure might not be able to detect different muscle properties among uniform groups (eg, subjects without symptoms). The data, therefore, might be best utilized in providing a reference range for a uniform group of subjects (eg, normative data). Further research is indicated to determine the utility of these measures in demonstrating differences between subjects with and without symptoms.

Conclusion

A new approach has been developed that allows a direct measure of EMG activity in the DCF muscles. The apparatus used in this approach consists of electrode contacts attached to a suction catheter, which are placed on the posterior oropharyngeal wall using a nasopharyngeal application. Further research is needed to investigate issues such as possible cross talk. The technique, however, looks promising for future examination of these DCF muscles in patients with neck pain to further understand the impairments that have been revealed in testing the CCF action in patients with whiplash-associated disorders and cervicogenic headache.

Table 1.
Mean, Standard Error of the Mean, and Ranges of the Normalized
Root-Mean-Square Values for the Deep Cervical Flexor (DCF),
Sternocleidomastoid (SCM), and Anterior Scalene (AS) Muscles Across the
5 Stages of the Craniocervical Flexion Test and of the Percentage of
Full Craniocervical Flexion Range of Motion (ROM) Used to Target Each
Pressure Level (N = 10)

            22 mm Hg
Muscle      X         SEM       Range

DCF (%)     37.97      4.00     19.76-60.88
SCM (%)     14.96      1.99      3.76-36.66
AS (%)      17.99      1.90      5.24-37.09
ROM (%)     17.87      9.46

            24 mm Hg
Muscle      X         SEM       Range

DCF (%)     45.46      3.48     30.31-63.89
SCM (%)     16.74      1.90      5.56-39.18
AS (%)      20.75      1.79      5.52-39.78
ROM (%)     34.98     13.73

            26 mm Hg
Muscle      X         SEM       Range

DCF (%)     53.32      3.81     32.28-73.24
SCM (%)     22.01      2.94      8.57-59.49
AS (%)      25.93      2.67      6.29-58.49
ROM (%)     55.78     10.74

            28 mm Hg
Muscle      X         SEM       Range

DCF (%)     62.91      4.27     37.09-82.22
SCM (%)     28.28      3.85     10.78-69.02
AS (%)      31.71      3.35     13.67-64.15
ROM (%)     71.20      9.40

            30 mm Hg
Muscle      X         SEM       Range

DCF (%)     77.63      5.46     41.87-98.77
SCM (%)     38.36      5.19     14.69-92.53
AS (%)      39.54      3.87     13.74-74.16
ROM (%)     88.43      6.60

Table 2.
Reliability of Normalized 1-Second Root-Mean-Square Values for the Left
Deep Cervical Flexor Muscles During the 5 Stages of the Craniocervical
Flexion (CCF) Test (a)

Stage of CCF          Variance Due            Variance Due
Test (mm Hg)          to Subjects             to Days

22                    21.8                    12.3
24                    14.1                    54.8
26                    39.3                    28.3
28                    63.1                    14.9
30                    59.8                    22.6

Variance Due          Within-Subject          Between-Subject
to Trials             nSEM                    nSEM

65.9                  10.3                     8.3
31.1                   8.9                     7.2
32.4                   8.8                     9.8
22.0                   7.0                    11.1
17.6                   6.7                    10.1

(a) Variance due to subjects based on intraclass correlation
coefficients; variance due to days and trials based on results of
analysis of variance. Normalized standard error of the mean (nSEM)
reflects deviation within and between subjects. Results are presented
as percentages of the total variance.

References

(1) Conley MS, Meyer RA, Bloomberg JJ, et al. Noninvasive analysis or human neck muscle function. Spine. 1995;20:2505-2512.

(2) Mayoux-Benhamou MA, Revel M, Vallee C, et al. Longus colli has a postural function on cervical curvature. Surg Radiol Anat. 1994;16: 367-371.

(3) Vasavada AN, Li S, Delp SL. Influence of muscle morphometry and moment arms on the moment-generating capacity of human neck muscles. Spine. 1998;23:412-422.

(4) Winters JM, Peles JD. Neck muscle activity and 3D head kinematics during quasistatic and dynamic tracking movements. In: Winters JM, Woo SLY, eds. Multiple Muscle Systems: Biomechanics and Movement Organization. New York, NY: Springer-Verlag; 1990:461-480.

(5) Uhlig Y, Weber BR, Grob D, Muntener M. Fiber composition and fiber transformations in neck muscles of patients with dysfunction of the cervical spine. J Orthop Res. 1995;13:240-249.

(6) Fountain FP, Minear WL, Allison RD. Function of longus colli and longissimus cervicis muscles in man. Arch Phys Med Rehabil. 1966;47: 665-669.

(7) Vitti M, Fujiwara M, Basmanjian JM, Iida M. The integrated roles of longus colli and sternocleidomastoid muscles: an electromyographic study. Anat Rec. 1973;177:471-484.

(8) Jull G, Barrett C, Magee R, Ho P. Further clinical clarification of the muscle dysfunction in cervical headache. Cephalalgia. 1999;19: 179-185.

(9) Jull GA. Deep cervical flexor muscle dysfunction in whiplash. Journal of Musculoskeletal Pain. 2000;8(1/2):143-154.

(10) Jull G, Trott P, Potter H, et al. A randomized controlled trial of exercise and manipulative therapy for cervicogenic headache. Spine. 2002;27:1835-1843.

(11) Monthly Index Medical Specialties. Monthly Index Medical Specialties (MIMS) Annual. Crows Nest, New South Wales, Australia: MediMedia; 2001.

(12) Hough A. Physiotherapy in Respiratory Care: A Problem Solving Approach to Respiratory and Cardiac Management. 2nd ed. London, England: Chapman Hill; 1996.

(13) Lang J. Clinical Anatomy of the Cervical Spine. New York, NY: Thieme Medical Publishers; 1993.

(14) Falla D, Dall'Alba P, Rainoldi A, et al. Location of innervation zones of sternocleidomastoid and scalene muscles: a basis fox clinical and research electromyography applications. Clin Neurophysiol 2092; 113: 57-63.

(15) Hermens HJ, Freriks B, Disselhorst-Klug C, Rau G. Development of recommendations for SEMG sensor and sensor placement procedures. J Electromyogr Kinesiol 2000;10:361-374.

(16) Falla D, Campbell C, Fagan A, et al. The relationship between upper cervical flexion range of motion and pressure change during the cranio-cervical flexion test. Man Ther. 2003;8:92-96.

(17) Pierce LNB. Guide to Mechanical Ventilation and Intensive Respiratory Care. Philadelphia, Pa: WB Saunders Co; 1995.

(18) Rainoldi A, Bullock-Saxton JE, Cavarretta F, Hogan N. Repeatability of maximal voluntary force and of surface EMG variables during voluntary isometric contraction of quadriceps muscles in healthy subjects. J Electromyogr Kinesiol. 2001;11:425-438.

(19) Rainoldi A, Galardi G, Maderna L, et al. Repeatability of surface EMG variables during voluntary isometric contractions of the biceps brachii muscle. J Electromyogr Kinesiol. 1999;9:105-119.

(20) McKenzie DK, Gandevia SC. Phrenic nerve conduction times and twitch pressures of the human diaphragm. J Appl Physiol. 1985;58: 1496-1504.

(21) Falla D, Dall'Alba P, Rainoldi A, et al. Repeatability of surface EMG variables in the sternocleidomastoid and anterior scalene muscles. Eur J Appl Physiol. 2002;87:542-549.

(*) Astra Pharmaceuticals, 50 Otis St, Westborough, MA 01581.

[dagger] Myotronics-Noromed Inc, 15425 53rd Ave S, Tukwila, WA 98188,

[double dagger] Laboratorio di Ingegneria del Sistema Neuromuscolare e della Riabilitazione Motoria, Via Cavalli 22G, Politecnico di Torino, Torino, Italy 10138.

[section] National Instruments Corp, 11500 N Mopac Expwy, Austin, TX 78759-3504.

[parallel] The MathWorks Inc, 3 Apple Hill Dr, Natick, MA 01760-2098.

(#) Chattanooga Group Inc, 4717 Adams Rd, Hixson, TN 37343.

(**) Amlab Technologies, 12 McGill St, Lewisham, New South Wales, Australia, 2049.

[double dagger] Cannon Australia Ply Ltd, 1 Thomas Holt Dr, North Ryde, New South Wales, Australia. 2113.

D Falla, PT, BPhty (Hons), is a doctoral student, Department of Physiotherapy, The University of Queensland, St Lucia, Brisbane, 4072 Queensland, Australia ([e-mail]). Address all correspondence to Ms Falla.

G Jull, PT, PhD, is Associate Professor and Head of Department, Department of Physiotherapy, The University of Queensland.

P Dall'Alba, PT, BPhty (Hons), is Research Officer, Department of Physiotherapy, The University of Queensland.

A Rainoldi is Physicist and a doctoral student, Centro di Bioingegneria, Dip di Elettronica, Politecnico di Torino, Italy, and Department of Physical Medicine and Rehabilitation, University of Tor Vergata and Fondazione Don Gnocchi, Rome, Italy.

R Merletti, PhD, is Engineer, Professor, and Director, Centro di Bioingegneria, Dip di Elettronica, Politecnico di Torino.

Ms Falla, Dr. Jull, and Mr Dall'Alba provided concept/research design. Ms Falla and Dr Jull provided writing. Ms Falla provided data collection. Ms Falla and Mr Rainoldi provided data analysis. Ms

Falla provided project management. Dr Jull provided facilities. Dr Jull and Dr Merletti provided fired procurement. Dr Jull, Mr Dall'Alba, Mr Rainoldi, and Dr Merletti provided consultation. The authors thank SOEM Medical SRL of Torino, Italy, for the suction unit provided during pilot studies.

Ethical approval for the study was granted by the Medical Research Ethics Committee of The University of Queensland, Australia.

This study was supported by a University of Queensland Small Grant and partially by the Fondazione CRT and Compagnia di San Patio di Torino.

This article was received November 5, 2002, and was accepted June 2, 2003.