- Email: [email protected]
Ultrasound assessment of the anatomical validity of T3 and L4 as sEMG recording sites
Journal of Biomechanics 44 (2011) 1025–1030
Contents lists available at ScienceDirect
Journal of Biomechanics journal homepage: www.elsevier.com/locate/jbiomech www.JBiomech.com
Ultrasound assessment of the anatomical validity of T3 and L4 as sEMG recording sites Diana M. Perriman a,b,n, Jennifer M. Scarvell a,b, Gordon S. Waddington c, Christian J. Lueck b,d, Andrew R. Hughes b,d, Teresa M. Neeman e, Paul N. Smith a,b a
Trauma and Orthopaedic Research Unit, Level 1, Building 6, Canberra Hospital, PO Box11, Woden, ACT 2606, Australia College of Medicine, Biology and Environment, The Australian National University, Canberra, ACT, Australia c Faculty of Health Sciences, University of Canberra, Canberra, ACT, Australia d Department of Neurology, Canberra Hospital, ACT, Australia e Statistical Consulting Unit, The Australian National University, Canberra, ACT, Australia b
a r t i c l e i n f o
abstract
Article history: Accepted 9 February 2011
The accuracy of surface EMG measurement is dependent upon minimizing potential crosstalk from other muscles. Although they are deeply situated, in places the erector spinae are covered with electrically silent aponeuroses rather than active muscle tissue. Theoretically these aponeuroses can serve as windows for sEMG recordings. A recent anatomical study concluded that T3 and L4 are ideal sites for recording the ES because the superficial muscle aponeuroses are wide at these sites. The aim of this prospective study was to investigate these sites in vivo using real time ultrasound. Ultrasound images from 20 subjects (10 o 30 years and 10 4 70 years; equal numbers of males and females in each group) were acquired during rest and in prone extension with the arms in three different positions. The most superficial aponeurosis widths were measured. The mean T3 aponeurosis width reduced significantly in extension from 4.4 7 4.7 mm at rest to 1.8 7 2.6 mm in extension (po 0.0001). Males had significantly smaller T3 aponeurosis widths than females (p¼ 0.049). The mean L4 aponeurosis width also significantly decreased in extension from 35.5 7 7.0 mm at rest to 29.9 7 7.2 mm in extension (p o 0.0001) due to ‘doming’ of the aponeurosis. Our results demonstrate that T3 is not a reliable site over which to record the ES because the aponeurosis width is too narrow. L4 is a good site if the electrodes are placed no more than 20 mm from the midline. & 2011 Elsevier Ltd. All rights reserved.
Keywords: Surface electromyography Ultrasound Erector spinae Trapezius aponeurosis Latissimus dorsi aponeurosis Thoracic spine Lumbar spine
1. Introduction Surface electromyography (sEMG) has been used extensively to measure muscular activity in the thoracic and lumbar erector spinae muscles (Table 1). It has advantages over needle electromyography (nEMG) in that it is non-invasive, painless and capable of measuring electrical activity from whole muscles rather than from single motor units (Meekins et al., 2008). This makes it an ideal tool for measuring muscle activity during functional activities in which nEMG needles might become dislodged or cause pain thereby inhibiting normal activation (Young et al., 1989; Soderberg and Knutson, 2000). Although sEMG has been used to estimate the activity of deeper muscles (McGill et al., 1996; Keenan et al., 2007), the EMG signal is dominated by the most superficial muscles (Fuglevand et al., 1992; Soderberg and Knutson, 2000) which will generate significant crosstalk
n
Corresponding author. Tel.: + 61 2 6244 4133; fax: + 61 2 6205 2157. E-mail address: [email protected] (D.M. Perriman).
0021-9290/$ - see front matter & 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.jbiomech.2011.02.007
(interference) unless they are completely inactive. (Soderberg and Knutson, 2000; Meekins et al., 2008). The erector spinae muscles (ES) are relatively deep. The thoraco-lumbar fascia overlies them along the entire length of the spine (Vleeming et al., 1995; Barker and Briggs, 1999). This fascia widens over the upper thoracic and lumbar regions where it acts as an aponeurosis connecting superficial muscles to the spinous processes (de Se ze and Cazalets, 2008). In the lumbar spine the fascia is thick and broad but in the thoracic spine it is thinner and narrower (Bogduk and Macintosh, 1984; Barker and Briggs, 1999). In theory, therefore, the widened areas of thoraco-lumbar fascia should provide several electrically silent windows for sEMG of the ES (de Se ze and Cazalets, 2008). Previous researchers have used numerous sites from which to record the ES (see Table 1), but they have done so without much consideration of these windows. In general, the selection of recording site has been made on the basis of prior use (Table 1). It is widely accepted that lumbar recording sites are relatively unaffected by extraneous signal from other muscles. However,
1026
D.M. Perriman et al. / Journal of Biomechanics 44 (2011) 1025–1030
Table 1 Examples of papers which have used surface EMG (sEMG) to evaluate the thoracic erector spinae (TES); the recording sites used; their justification for the sites chosen and the main conclusion with respect to their thoracic recordings. The citation trail leads either nowhere, or to Lafortune et al. whose findings, which were published only in conference proceedings, were speculative. All measurements are expressed in mm for consistency. Paper
Erector spinae sEMG electrode placement
Justification for electrode site
Main conclusions from thoracic recording
Floyd and Silver (1951)
Lateral to T10, T12, L2 and L4
Full spinal flexion leads to relaxation of all the erector spinae muscles
Schultz et al. (1982)
30 mm lat to T8, 20 mm lat to L3 50 mm lat to T9 Lat to L2 and above PSIS 30 mm lat to L3 50 mm lat to T9 30 mm lat to L3
‘we have some evidence to show that the upper part of the muscle in the thoracic region behaves the same way as the lumbar part’—not specified None
Lafortune et al. (1988)
McGill (1991)
McGill (1992) Cholewicki and McGill (1996) Seelen et al. (1998) Krajcarski et al. (1999) Callaghan and Dunk (2002) Moseley et al. (2003) Abdoli-E et al. (2006)
50 mm 30 mm 50 mm 30 mm 25 mm 25 mm 40 mm 30 mm 50 mm 30 mm 50 mm
lat lat lat lat lat lat T9 L3 T9 lat lat
to to to to to to
T9 L3 T9 L3 T3 and T9 L3
Recordings over T9 and L3 site show insignificant differences in 8 young males during a lifting task Lafortune et al. (1988)
None None None McGill (1992) McGill (1991)
L3 T7
None
50 mm T9 30 mm L4 50 mm lat to T9
None
Greig et al. (2008)
Lateral to T8 Lateral to L3
Schultz et al. (1982)
Cholewicki et al. (2009)
50 mm lat to T9
Cholewicki and McGill (1996)
O’Sullivan et al. (2006)
Callaghan and Dunk (2002)
there is less certainty about thoracic sites. Floyd and Silver (1955) were the first to study ES using sEMG (Floyd and Silver, 1955). They positioned their recording electrodes adjacent to the spine at various locations and noted that arms and shoulders must be inactivated by hanging down loosely under gravity in order to ensure no crosstalk from trapezius and latissimus dorsi muscles. In a later study, Lafortune et al. (1988) determined that the thoracic ES could be faithfully recorded 40 mm lateral to T9 because this site yielded the same sEMG amplitudes as a site 30 mm lateral to L4 (Lafortune et al., 1988). This ‘evidence’ has subsequently been widely cited (see Table 1). Nevertheless, concern that the ES recording sites had not been adequately validated prompted a recent anatomical study by de Se ze and Cazalets who sought to identify the optimal ES recording sites for sEMG using cadaveric dissection (de Se ze and Cazalets, 2008). They found fascial windows through which the ES could theoretically be recorded at C7, T3, and T12 to L4, with T3 and L4 being the most reliable sites. The site adjacent to T9 was found to be covered by the trapezius and or latissimus dorsi muscles in all cases. The sites identified by de Se ze and Cazalets were, however, identified on cadavers and it remains possible that the extent of the fascia is different in live subjects for a number of reasons. First, cadaveric muscle architecture differs from living muscle architecture: aged and preserved tissues can be flattened and thinned making it difficult to find margins between contractile and non-contractile tissues (Salmhofer et al., 1996; Martin et al., 2001). Second, the resting position of muscle after death may be different to that during life (Tsui et al., 2008). Third, stretching a muscle by changing the position of the arms or shoulders may
Did not really use the results—‘similar to L3’ sEMG recordings 30 mm lat to L3 may represent activity in upper paraspinal muscles Thoracic (and lumbar) ES perform a balancing and stabilizing role while other muscles generate axial torque During torsion the lumbar and thoracic erector spinae behave similarly EMG assisted algorithm suggests low load tasks may cause lumbar spine buckling. Low thoracic SCI patients recruit LES and TES for sitting balance Thoracic and lumbar ES stabilize the trunk better after preloading Slumped sitting led to flexion relaxation in TES but not LES TES switch on prior to multifidus in response to perturbation Lift device reduced activity in TES TES more active in ‘thoracic’ sitting therefore produce higher compressive loads and fatigue more quickly No change in thoracic sEMG with change in walking surface whereas there was in the lumbar spine Sig. diff in TES activity during static traction compared to sinusoidal traction
affect the relative extent of the fascia. Finally, the cadavers used in de Se ze and Cazalets0 study were presumably elderly which may limit the application of their findings to the elderly. Accordingly, the aim of this study was to measure the width of the fascial windows (aponeuroses) in live subjects using real-time ultrasound (US) imaging. On the basis of de Se ze and Cazelets’ findings (de Se ze and Cazalets, 2008), the aponeuroses at T3 and L4 were selected as potential sites for sEMG of the thoracic and lumbar ES, respectively. Measurements were made on a group of younger ( o30 years) and a group of older ( 470 years) participants to assess the effect of aging.
2. Methods A power analysis based on de Se ze and Cazelets0 results indicated that ten subjects in each age group would be sufficient to address our hypotheses. Accordingly, 20 healthy volunteers were recruited for this study by advertisement in a hospital. Ten subjects were aged between 18 and 30 years and ten were over 70 years, with equal numbers of males and females in each group (Table 2). Exclusion criteria included previous significant spinal problems such as fracture, surgery or disease; thoracotomy; pregnancy, neuropathy or myopathy. All subjects gave informed consent. Ethics approval for this study was granted by the ACT Health Human Research Ethics Committee and the Australian National University Human Research Ethics Committee. Ultrasonography was performed by DP or JS under the guidance of a trained sonographer using a Mindray DP-6600 ultrasound scanner with a 3.5 MHz R50 electronic convex array transducer (Mindray Biomedical, Shenzhen, China). The participants were positioned in prone lying. After palpating all of the spinous processes, the T3 and L4 spinous processes were marked with ink. The US transducer head was always positioned to the right of the T3 and L4 spinous processes because aponeurosis widths have previously been reported to be symmetrical (de Se ze and Cazalets, 2008). US images were acquired in each of
D.M. Perriman et al. / Journal of Biomechanics 44 (2011) 1025–1030
Table 2 The subject demographics, mean 7 standard deviation and [range]. The younger (o 30) and the older (470) groups each included 5 males and 5 females. Age category
Age (years)
Weight (kg)
Height (cm)
o30 470
23.9 7 1.8 [21–27] 76.1 7 4.6 [70–84]
68.3 7 10.8 [55–91] 73.2 7 8.7 [57–88]
173.6 711.1 [157–190] 173.3 711.5 [161–197]
Fig. 1. US images of the T3 spinous process (SP) with the attachments of the right trapezius (Tr), rhomboids (Rh) and the erector spinae (ES): (A) at rest and (B) in extension. Arrows indicate the points between which the aponeurosis width measurement was made. Three layers of fascia are identified: the trapezius aponeurosis (a), the rhomboid aponeurosis which is continuous with the superficial lamina of the posterior layer of the thoraco-lumbar fascia (b), and the deep lamina of the posterior layer of the thoraco-lumbar fascia (c). The trapezius aponeurosis was seen to be separate from the rhomboid aponeurosis during extension. three positions: (i) arms above the head (‘above head’), (ii) arms hanging by the side (‘hanging’) and (iii) arms lying alongside the body (‘by side’). These positions were chosen in order to examine the effect of changes in superficial muscle position on the T3 aponeurosis width. The order in which the subjects assumed
1027
each position was randomized and the locations of the spinous processes were reassessed after each change of position. US images were acquired at rest and then again during sustained back extension from prone for each of the three positions, resulting in a total of six thoracic images for each subject. Six lumbar images were then recorded in the same way. Measurements of aponeurosis width were made offline from the US images. At T3, aponeurosis width measurements were made by identifying the hyperechoic fascial lines situated above and below the trapezius muscle and measuring the distance between the point where they coalesced (the medial edge of the muscle) and the tip of the spinous process (Fig. 1A and B). L4 aponeurosis width measurements were made by identifying the fascial lines above and below the latissimus dorsi and measuring the distance between the point just lateral to where they coalesced and the tip of the spinous process (Fig. 2A and B). If an image was insufficiently clear to delineate the precise point at which the fascia coalesced, a conservative judgment in favor of a greater aponeurosis width was
Fig. 2. US images of the L4 spinous process (SP) with the attachments of the right latisimus dorsi (LD) and the erector spinae (ES). (A) at rest and (B) in extension. Arrows indicate the points between which the aponeurosis width measurement was made. Two layers of fascia are identified: the latisimus dorsi aponeurosis which is continuous with the superficial lamina of the posterior layer of the thoraco-lumbar fascia (a), and the deep lamina of the posterior layer of the thoraco-lumbar fascia (b). The two layers were pushed upwards during extension into a domed shape due to the contraction of the ES resulting in a narrowing of the aponeurotic window.
1028
D.M. Perriman et al. / Journal of Biomechanics 44 (2011) 1025–1030
Table 3 Measurements of aponeurosis width (mm) for the whole group, those o 30 years of age (n¼ 10) and 470 years (n ¼10) at T3 and L4. The measurements at rest and during extension in prone are presented separately.
T3 Rest Mean 7 SD Range Extended Mean 7 SD Range L4 Rest Mean 7 SD Range Extended Mean 7 SD Range
All subjects
o 30 years
Females and males
Female
Male
Female
Male
4.4 7 4.7 0–21.2
5.6 75.8 0–21.2
0.37 0.7 0–2.3
5.9 73.8 0–10.7
5.6 7 4.6 0–17.4
1.8 7 2.6 0–9.1
2.3 72.3 0–8.2
0.27 0.4 0–1.2
3.4 73.3 0–9.1
1.4 7 2.2 0–5.9
35.5 7 7.0 18–52.8
36.8 7 4.6 28.1–44.9
40.47 4.8 29.8–48.7
30.3 75.2 18–38.8
34.7 7 8.9 25.2–52.8
29.4 7 7.1 14–52.6
25.9 7 8.4 19.6–35.3
33.17 5.4 24.9–42.9
25.5 7 6.1 14–33.2
31.6 7 9.5 22.8–52.6
470 years
made. All measurements were made using the linear measurement software incorporated within the Mindray system (Figs. 1 and 2).
>70
<30 50
The data for each group were expressed as mean and standard deviations (mean (7 SD)). A split-plot ANOVA model with age group and gender as between-subject effects, and activity (at rest and in extension), arm position and arm position order as within-subject effects, was used for both thoracic and lumbar data. Differences were considered significant when p o0.05.
aponeurosis width (mm)
2.1. Statistical analysis
40 30 20 10
3. Results 3.1. Thoracic (T3) The mean aponeurosis width between the tip of the spinous process of T3 and the adjacent trapezius muscle was significantly smaller during extension than at rest (1.872.6 mm vs. 4.474.7 mm, p o0.0001, Table 3). Neither arm position nor the order in which the positions were assumed had a significant effect. Males had significantly smaller mean aponeurosis widths than females (p¼0.049, Table 3 and Fig. 3). This difference was most marked in young males who had measurements approaching zero during both rest and extension, contrasting with young females and both males and females in the older group, all of whom had mean aponeurosis widths of between 5.6 and 5.9 mm at rest (Table 3 and Fig. 3). 3.2. Lumbar (L4) The mean aponeurosis width between the tip of the spinous process of L4 and the medial edge of the adjacent latissimus dorsi muscle was significantly smaller during extension than at rest (29.477.1 mm vs. 35.5 77.0 mm, po0.0001, Fig. 3, Table 3). The order in which the positions were assumed did not have a significant effect. 3.3. The dynamic behavior of the muscle with respect to the aponeurosis The appearance and behavior of the aponeuroses in the thoracic and lumbar regions were very different. The most superficial muscle at the thoracic site was the trapezius. At rest, this muscle was attached to the T3 spinous process via a clearly visible
0 female
male
female
Thoracic Rest
Thoracic Extension
Lumbar Rest
Lumbar Extension
male
Fig. 3. Boxplot of the aponeurosis widths at T3 and L4 during rest and extension for the males and females in both the younger and older age groups. The aponeurosis width at rest was significantly wider than in extension at both sites (p o 0.0001). Males had significantly narrower T3 aponeurosis widths than females (p o 0.05).
aponeurosis in 15 out of the 20 subjects. In the remaining five subjects the muscle inserted directly onto the spinous process with no discernable aponeurosis. Where an aponeurosis was present, its width decreased during active extension. In three of the 15 subjects who had a discernable aponeurosis at rest, it was reduced to nothing during extension. When present, the aponeurosis appeared to connect both the trapezius and the rhomboid muscles via a single structure at rest, but during extension it was seen to cleave horizontally as the trapezius muscle fibers thickened within the space revealing separate attachments. In the five subjects whose trapezius appeared to insert directly onto the spinous process, this thickening was profound. The rhomboid muscle, which was positioned deep to the trapezius, increased in thickness but always inserted onto the spinous process via an aponeurosis situated below the trapezius muscle or the trapezius aponeurosis (Fig. 1B). In contrast to the T3 aponeurosis, the L4 aponeurosis was wide and thick and did not appear to cleave during extension but was seen to bulge or dome posteriorly over the thickened ES (Fig. 2B). The latissimus dorsi muscle belly was seen to thicken, but its relation to the aponeurosis did not differ markedly from that seen in the resting position (Fig. 2A and B).
D.M. Perriman et al. / Journal of Biomechanics 44 (2011) 1025–1030
4. Discussion The aim of this study was to investigate the utility of two recently proposed sEMG recording sites for the ES in vivo. There were four major findings. First, the most superficial aponeurosis width adjacent to T3 was either non-existent or too narrow to allow sEMG recording. Second, aponeurosis widths at both T3 and L4 reduced significantly during active back extension. Third, aponeurosis width at T3 was significantly narrower in men than in women with young men often demonstrating no discernable aponeurosis at all. Finally, the trapezius muscle attached to the spinous process at T3 within an envelope of connective tissue which was separate to the aponeurosis of the rhomboid muscle lying below it. These findings have not previously been reported in the literature. The accuracy of direct recording of muscle sEMG amplitudes depends critically on the positioning of the recording electrodes so as to avoid crosstalk from other muscles (Soderberg and Knutson, 2000). The ES have been studied using sEMG for nearly 60 years (Table 1). Surface EMG recording electrodes for the thoracic ES are most commonly sited in the lower thoracic spine because of concerns over crosstalk from the trapezius and rhomboid muscles (Floyd and Silver, 1951). The most common site used is 40–50 mm lateral to T9 (McGill, 1991; Callaghan and Dunk, 2002). This site was first suggested by Lafortune et al. (1988) who found that sEMG signal amplitudes measured here and 30 mm adjacent to L4 were the same. These recordings were made from eight young men while they performed squat lifts (Lafortune et al., 1988). Until now this site has been the accepted standard for electrode placement. As noted previously, recent anatomical evidence has suggested that recordings made adjacent to T9 are likely to be subject to significant crosstalk (de Se ze and Cazalets, 2008). Cadaveric dissection revealed that the thoracic ES were covered by superficial muscle tissue from T5 to T11 but de Se ze and Cazalets suggested that there was an ‘electrically silent window’ between C7 and T4 through which the thoracic ES could potentially be recorded. This window was widest at T3 with a mean width of 34 mm. They recommended positioning sEMG electrodes 20 mm from the midline at T3. Other investigators have located their electrodes 25 mm adjacent to the T3 spinous process (Seelen et al., 1998). Our in vivo study demonstrated, however, that the mean T3 aponeurosis width at rest was only 6 mm reducing to 3 mm during active back extension. Thus, according to our results the sites suggested by de Se ze and Cazalets and those used by previous researchers would lie over the belly of the trapezius muscle. The T3 site is therefore not suitable for sEMG recording of the thoracic ES in vivo, especially during functional activities, unless the trapezius and rhomboid muscles are specifically rendered completely inactive. In the light of both our results and those of de Se ze and Cazalets the findings of previous studies which have used sEMG to record the thoracic ES may need reconsideration (Table 1). Although difficult to obtain directly, the activity of deeply situated muscles can be inferred from sEMG. McGill et al. (1996) hypothesized that deep trunk muscle activity could be predicted from surface activity because the neural drive to both is matched. They found that the difference in amplitude between superficial and deep recordings was generally within 15% and argued that this magnitude of error was within acceptable limits for biomechanical modeling (McGill et al., 1996). Whether there is a uniform relationship between the surface muscles and the deeper thoracic ES has not yet been established. The suitability of L4 as a recording site is not in dispute and our findings confirm that the aponeurosis adjacent to L4 is sufficiently wide for direct sEMG recording. de Se ze and Cazalets
1029
recorded a mean aponeurosis width of 86 mm and recommended that sEMG electrodes be sited at a distance of 40 mm from the midline at L4 (de Se ze and Cazalets, 2008). Many previous studies have used distances of between 20 and 30 mm (Table 1). Importantly, our study showed that the mean aponeurosis width reduced significantly in extension from 35.5 mm to just 29 mm. This reduction in aponeurosis width was the result of a ‘doming’ effect produced by the contraction of the underlying ES during active extension (Fig. 2B). In fact, this ‘doming’ can easily be observed from the skin surface but this is the first time that the effect on the aponeurotic window has been reported. Our results indicate that if subjects are to be studied during extension a distance of 20 mm from the L4 spinous process would ensure that the electrodes were positioned over the aponeurosis, thereby preventing crosstalk from the latissimus dorsi. The finding that the lumbar aponeurosis width may be reduced during strong contraction of the erector spinae is relevant to the findings of Lafortune et al. (1988). As previously noted, their assertion that the T9 site records the thoracic ES rests on the observation that lumbar ES signal amplitudes measured at 30 mm adjacent to L4 were the same as signal amplitudes measured 40 mm lateral to T9. In our study the mean aponeurosis width at L4 in young males during extension was 33 mm so siting the recording electrode at 30 mm may have resulted in significant crosstalk from the latissimus dorsi in their study. Further, the site adjacent to T9 is completely covered by the muscle fibers of either the trapezius or latissimus dorsi (de Se ze and Cazalets, 2008) raising the strong possibility that the signal detected by Lafortune et al. may actually have arisen from the latissimus dorsi which would have been strongly activated during the squat-lift protocol used in their study. The US images of the aponeurosis showed that the rhomboid and trapezius muscles were surrounded by separate fascial layers which, in most cases, coalesced to form their aponeurotic attachments onto the spinous process. For trapezius the extension task demonstrated this arrangement very clearly: the muscle expanded within its fascial envelope during contraction, dramatically reducing the fascial window and removing the possibility of any ‘electrically silent’ access to the ES muscles below. This configuration is consistent with descriptions of the trapezius lying between two laminae of the deep cervical fascia (Newell, 2008). The rhomboid aponeurosis did not appear to reduce in length during the extension task. The images of the lumbar aponeurosis were consistent with previous anatomical descriptions (Bogduk and Macintosh, 1984; Vleeming et al., 1996; Barker and Briggs, 1999; Loukas et al., 2008). The limitations of this study include small sample size and limited US quality. Though small, the number of subjects used in this study was sufficient to address its main aim. A larger cohort might further clarify the age and gender effects which we observed. A 3.5 MHz transducer was used to acquire the images for this study. A higher frequency US unit would undoubtedly have resulted in more detailed images (Salmhofer et al., 1996). To overcome this potential limitation, we were conservative in our measurements in cases where clarity of image was an issue, reporting larger width measurements in all cases. We are therefore confident that the results do not underestimate aponeurosis width. Finally, because the study participants were only positioned in prone lying extrapolating our findings to other positions should be done with caution.
5. Conclusion This study reveals the potential risks of extrapolating findings from cadaveric studies to living systems. It transpires that T3 is
1030
D.M. Perriman et al. / Journal of Biomechanics 44 (2011) 1025–1030
not a reliable site over which to record the ES directly in vivo because the trapezius muscle, when contracted, covers the ES. This is consistent with the suggestion that many previous studies which have purported to measure the thoracic ES using sEMG have, in fact, recorded activity from the superficial muscles. In the lumbar spine we would recommend electrode placement no further than 20 mm adjacent to L4. On the basis of our results, future researchers are encouraged to re-examine the use of sEMG to study the ES. In theory, nEMG would be more accurate but the difficulties associated with its use make it less practical. An alternative approach would be to use sEMG and compensate for crosstalk from superficial muscles by employing techniques which use a larger number of recording sites along with more advanced signal processing analysis.
Conflict of interest statement None of the authors have any financial and personal relationships with other people or organizations that could have inappropriately influenced this work including employment, consultancies, stock ownership, honoraria, paid expert testimony and patent applications/registrations. Funding for this work was provided by an NH&MRC PhD scholarship, the Canberra Hospital Private Practice Fund and AO Spine Pacific.
Acknowledgments The authors would like to thank the study participants, Roxanne Miller and Annette Lawson for their invaluable assistance with this study, and also Dr. Nick Ball for his very helpful advice. We gratefully acknowledge the support of the NHMRC Australia, AO Spine Australasia and the Canberra Hospital Private Practice Fund. We would also like to respectfully acknowledge the anonymous reviewers who greatly enhanced this manuscript with their thoughtful suggestions. References Abdoli-E, M., Agnew, M.J., et al., 2006. An on-body personal lift augmentation device (PLAD) reduces EMG amplitude of erector spinae during lifting tasks. Clinical Biomechanics 21 (5), 456–465. Barker, P.J., Briggs, C.A., 1999. Attachments of the posterior layer of lumbar fascia. Spine 24 (17), 1757–1764. Bogduk, N., Macintosh, J.E., 1984. The applied anatomy of the thoracolumbar fascia. Spine 9 (2), 164–170. Callaghan, J.P., Dunk, N.M., 2002. Examination of the flexion relaxation phenomenon in erector spinae muscles during short duration slumped sitting. Clinical Biomechanics (Bristol, Avon) 17 (5), 353–360. Cholewicki, J., Lee, A.S., et al., 2009. Trunk muscle response to various protocols of lumbar traction. Manual Therapy 14 (5), 562–566. Cholewicki, J., McGill, S.M., 1996. Mechanical stability of the in vivo lumbar spine: implications for injury and chronic low back pain. Clinical Biomechanics (Bristol, Avon) 11 (1), 1–15. de Se ze, M.P., Cazalets, J.R., 2008. Anatomical optimization of skin electrode placement to record electromyographic activity of erector spinae muscles. Surgical and Radiolic Anatomy 30 (2), 137–143.
Floyd, W.F., Silver, P.H., 1951. Function of erectores spinae in flexion of the trunk. Lancet 1 (6647), 133–134. Floyd, W.F., Silver, P.H., 1955. The function of the erectores spinae muscles in certain movements and postures in man. Journal of Physiology 129 (1), 184–203. Fuglevand, A.J., Winter, D.A., et al., 1992. Detection of motor unit action potentials with surface electrodes: influence of electrode size and spacing. Biological Cybernetics 67 (2), 143–153. Greig, A.M., Bennell, K.L., et al., 2008. Postural taping decreases thoracic kyphosis but does not influence trunk muscle electromyographic activity or balance in women with osteoporosis. Manual Therapy 13 (3), 249–257. Keenan, K.G., Farina, D., et al., 2007. Sensitivity of the cross-correlation between simulated surface EMGs for two muscles to detect motor unit synchronization. Journal of Applied Physiology 102 (3), 1193–1201. Krajcarski, S.R., Potvin, J.R., et al., 1999. The in vivo dynamic response of the spine to perturbations causing rapid flexion: effects of pre-load and step input magnitude. Clinical Biomechanics (Bristol, Avon) 14 (1), 54–62. Lafortune, D., Norman, R., et al., 1988. Ensemble averages linear enveloped EMGs during lifting. In: Proceedings of the Biannual Conference of the Canadian Society for Biomechanics, Ottawa, Canada. Loukas, M., Shoja, M.M., et al., 2008. Anatomy and biomechanics of the vertebral aponeurosis part of the posterior layer of the thoracolumbar fascia. Surgical and Radiolic Anatomy 30 (2), 125–129. Martin, D.C., Medri, M.K., et al., 2001. Comparing human skeletal muscle architectural parameters of cadavers with in vivo ultrasonographic measurements. Journal of Anatomy 199 (Part 4), 429–434. McGill, S., Juker, D., et al., 1996. Appropriately placed surface EMG electrodes reflect deep muscle activity (psoas, quadratus lumborum, abdominal wall) in the lumbar spine. Journal of Biomechanics 29 (11), 1503–1507. McGill, S.M., 1991. Electromyographic activity of the abdominal and low back musculature during the generation of isometric and dynamic axial trunk torque: implications for lumbar mechanics. Journal of Orthopedic Research 9 (1), 91–103. McGill, S.M., 1992. The influence of lordosis on axial trunk torque and trunk muscle myoelectric activity. Spine 17 (10), 1187–1193. Meekins, G.D., So, Y., et al., 2008. American Association of Neuromuscular and Electrodiagnostic Medicine evidenced-based review: use of surface electromyography in the diagnosis and study of neuromuscular disorders. Muscle and Nerve 38 (4), 1219–1224. Moseley, G.L., Hodges, P.W., et al., 2003. External perturbation of the trunk in standing humans differentially activates components of the medial back muscles. Journal of Physiology 547 (Part 2), 581–587. Newell, L., 2008. Section 5—the back. In: Standring, S. (Ed.), Gray’s Anatomy 40th Edn. Churchill Livingstone, New York, pp. 705–773. O0 Sullivan, P.B., Dankaerts, W., et al., 2006. Effect of different upright sitting postures on spinal-pelvic curvature and trunk muscle activation in a pain-free population. Spine 31 (19), E707–E712. Salmhofer, W., Rieger, E., et al., 1996. Influence of skin tension and formalin fixation on sonographic measurement of tumor thickness. Journal of the American Academy of Dermatology 34 (1), 34–39. Schultz, A., Andersson, G.B., et al., 1982. Analysis and quantitative myoelectric measurements of loads on the lumbar spine when holding weights in standing postures. Spine 7 (4), 390–397. Seelen, H.A., Potten, Y.J., et al., 1998. Development of new muscle synergies in postural control in spinal cord injured subjects. Journal of Electromyography and Kinesiology 8 (1), 23–34. Soderberg, G.L., Knutson, L.M., 2000. A guide for use and interpretation of kinesiologic electromyographic data. Physical Therapy 80 (5), 485–498. Tsui, B., Dillane, D., et al., 2008. Ultrasound imaging in cadavers: training in imaging for regional blockade at the trunk. Canadian Journal of Anesthesia 55 (2), 105–111. Vleeming, A., Pool-Goudzwaard, A.L., et al., 1996. The function of the long dorsal sacroiliac ligament: its implication for understanding low back pain. Spine 21 (5), 556–562. Vleeming, A., Pool-Goudzwaard, A.L., et al., 1995. The posterior layer of the thoracolumbar fascia. Its function in load transfer from spine to legs. Spine 20 (7), 753–758. Young, C.C., Rose, S.E., et al., 1989. The effect of surface and internal electrodes on the gait of children with cerebral palsy, spastic diplegic type. Journal of Orthopedic Research 7 (5), 732–737.
Contents lists available at ScienceDirect
Journal of Biomechanics journal homepage: www.elsevier.com/locate/jbiomech www.JBiomech.com
Ultrasound assessment of the anatomical validity of T3 and L4 as sEMG recording sites Diana M. Perriman a,b,n, Jennifer M. Scarvell a,b, Gordon S. Waddington c, Christian J. Lueck b,d, Andrew R. Hughes b,d, Teresa M. Neeman e, Paul N. Smith a,b a
Trauma and Orthopaedic Research Unit, Level 1, Building 6, Canberra Hospital, PO Box11, Woden, ACT 2606, Australia College of Medicine, Biology and Environment, The Australian National University, Canberra, ACT, Australia c Faculty of Health Sciences, University of Canberra, Canberra, ACT, Australia d Department of Neurology, Canberra Hospital, ACT, Australia e Statistical Consulting Unit, The Australian National University, Canberra, ACT, Australia b
a r t i c l e i n f o
abstract
Article history: Accepted 9 February 2011
The accuracy of surface EMG measurement is dependent upon minimizing potential crosstalk from other muscles. Although they are deeply situated, in places the erector spinae are covered with electrically silent aponeuroses rather than active muscle tissue. Theoretically these aponeuroses can serve as windows for sEMG recordings. A recent anatomical study concluded that T3 and L4 are ideal sites for recording the ES because the superficial muscle aponeuroses are wide at these sites. The aim of this prospective study was to investigate these sites in vivo using real time ultrasound. Ultrasound images from 20 subjects (10 o 30 years and 10 4 70 years; equal numbers of males and females in each group) were acquired during rest and in prone extension with the arms in three different positions. The most superficial aponeurosis widths were measured. The mean T3 aponeurosis width reduced significantly in extension from 4.4 7 4.7 mm at rest to 1.8 7 2.6 mm in extension (po 0.0001). Males had significantly smaller T3 aponeurosis widths than females (p¼ 0.049). The mean L4 aponeurosis width also significantly decreased in extension from 35.5 7 7.0 mm at rest to 29.9 7 7.2 mm in extension (p o 0.0001) due to ‘doming’ of the aponeurosis. Our results demonstrate that T3 is not a reliable site over which to record the ES because the aponeurosis width is too narrow. L4 is a good site if the electrodes are placed no more than 20 mm from the midline. & 2011 Elsevier Ltd. All rights reserved.
Keywords: Surface electromyography Ultrasound Erector spinae Trapezius aponeurosis Latissimus dorsi aponeurosis Thoracic spine Lumbar spine
1. Introduction Surface electromyography (sEMG) has been used extensively to measure muscular activity in the thoracic and lumbar erector spinae muscles (Table 1). It has advantages over needle electromyography (nEMG) in that it is non-invasive, painless and capable of measuring electrical activity from whole muscles rather than from single motor units (Meekins et al., 2008). This makes it an ideal tool for measuring muscle activity during functional activities in which nEMG needles might become dislodged or cause pain thereby inhibiting normal activation (Young et al., 1989; Soderberg and Knutson, 2000). Although sEMG has been used to estimate the activity of deeper muscles (McGill et al., 1996; Keenan et al., 2007), the EMG signal is dominated by the most superficial muscles (Fuglevand et al., 1992; Soderberg and Knutson, 2000) which will generate significant crosstalk
n
Corresponding author. Tel.: + 61 2 6244 4133; fax: + 61 2 6205 2157. E-mail address: [email protected] (D.M. Perriman).
0021-9290/$ - see front matter & 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.jbiomech.2011.02.007
(interference) unless they are completely inactive. (Soderberg and Knutson, 2000; Meekins et al., 2008). The erector spinae muscles (ES) are relatively deep. The thoraco-lumbar fascia overlies them along the entire length of the spine (Vleeming et al., 1995; Barker and Briggs, 1999). This fascia widens over the upper thoracic and lumbar regions where it acts as an aponeurosis connecting superficial muscles to the spinous processes (de Se ze and Cazalets, 2008). In the lumbar spine the fascia is thick and broad but in the thoracic spine it is thinner and narrower (Bogduk and Macintosh, 1984; Barker and Briggs, 1999). In theory, therefore, the widened areas of thoraco-lumbar fascia should provide several electrically silent windows for sEMG of the ES (de Se ze and Cazalets, 2008). Previous researchers have used numerous sites from which to record the ES (see Table 1), but they have done so without much consideration of these windows. In general, the selection of recording site has been made on the basis of prior use (Table 1). It is widely accepted that lumbar recording sites are relatively unaffected by extraneous signal from other muscles. However,
1026
D.M. Perriman et al. / Journal of Biomechanics 44 (2011) 1025–1030
Table 1 Examples of papers which have used surface EMG (sEMG) to evaluate the thoracic erector spinae (TES); the recording sites used; their justification for the sites chosen and the main conclusion with respect to their thoracic recordings. The citation trail leads either nowhere, or to Lafortune et al. whose findings, which were published only in conference proceedings, were speculative. All measurements are expressed in mm for consistency. Paper
Erector spinae sEMG electrode placement
Justification for electrode site
Main conclusions from thoracic recording
Floyd and Silver (1951)
Lateral to T10, T12, L2 and L4
Full spinal flexion leads to relaxation of all the erector spinae muscles
Schultz et al. (1982)
30 mm lat to T8, 20 mm lat to L3 50 mm lat to T9 Lat to L2 and above PSIS 30 mm lat to L3 50 mm lat to T9 30 mm lat to L3
‘we have some evidence to show that the upper part of the muscle in the thoracic region behaves the same way as the lumbar part’—not specified None
Lafortune et al. (1988)
McGill (1991)
McGill (1992) Cholewicki and McGill (1996) Seelen et al. (1998) Krajcarski et al. (1999) Callaghan and Dunk (2002) Moseley et al. (2003) Abdoli-E et al. (2006)
50 mm 30 mm 50 mm 30 mm 25 mm 25 mm 40 mm 30 mm 50 mm 30 mm 50 mm
lat lat lat lat lat lat T9 L3 T9 lat lat
to to to to to to
T9 L3 T9 L3 T3 and T9 L3
Recordings over T9 and L3 site show insignificant differences in 8 young males during a lifting task Lafortune et al. (1988)
None None None McGill (1992) McGill (1991)
L3 T7
None
50 mm T9 30 mm L4 50 mm lat to T9
None
Greig et al. (2008)
Lateral to T8 Lateral to L3
Schultz et al. (1982)
Cholewicki et al. (2009)
50 mm lat to T9
Cholewicki and McGill (1996)
O’Sullivan et al. (2006)
Callaghan and Dunk (2002)
there is less certainty about thoracic sites. Floyd and Silver (1955) were the first to study ES using sEMG (Floyd and Silver, 1955). They positioned their recording electrodes adjacent to the spine at various locations and noted that arms and shoulders must be inactivated by hanging down loosely under gravity in order to ensure no crosstalk from trapezius and latissimus dorsi muscles. In a later study, Lafortune et al. (1988) determined that the thoracic ES could be faithfully recorded 40 mm lateral to T9 because this site yielded the same sEMG amplitudes as a site 30 mm lateral to L4 (Lafortune et al., 1988). This ‘evidence’ has subsequently been widely cited (see Table 1). Nevertheless, concern that the ES recording sites had not been adequately validated prompted a recent anatomical study by de Se ze and Cazalets who sought to identify the optimal ES recording sites for sEMG using cadaveric dissection (de Se ze and Cazalets, 2008). They found fascial windows through which the ES could theoretically be recorded at C7, T3, and T12 to L4, with T3 and L4 being the most reliable sites. The site adjacent to T9 was found to be covered by the trapezius and or latissimus dorsi muscles in all cases. The sites identified by de Se ze and Cazalets were, however, identified on cadavers and it remains possible that the extent of the fascia is different in live subjects for a number of reasons. First, cadaveric muscle architecture differs from living muscle architecture: aged and preserved tissues can be flattened and thinned making it difficult to find margins between contractile and non-contractile tissues (Salmhofer et al., 1996; Martin et al., 2001). Second, the resting position of muscle after death may be different to that during life (Tsui et al., 2008). Third, stretching a muscle by changing the position of the arms or shoulders may
Did not really use the results—‘similar to L3’ sEMG recordings 30 mm lat to L3 may represent activity in upper paraspinal muscles Thoracic (and lumbar) ES perform a balancing and stabilizing role while other muscles generate axial torque During torsion the lumbar and thoracic erector spinae behave similarly EMG assisted algorithm suggests low load tasks may cause lumbar spine buckling. Low thoracic SCI patients recruit LES and TES for sitting balance Thoracic and lumbar ES stabilize the trunk better after preloading Slumped sitting led to flexion relaxation in TES but not LES TES switch on prior to multifidus in response to perturbation Lift device reduced activity in TES TES more active in ‘thoracic’ sitting therefore produce higher compressive loads and fatigue more quickly No change in thoracic sEMG with change in walking surface whereas there was in the lumbar spine Sig. diff in TES activity during static traction compared to sinusoidal traction
affect the relative extent of the fascia. Finally, the cadavers used in de Se ze and Cazalets0 study were presumably elderly which may limit the application of their findings to the elderly. Accordingly, the aim of this study was to measure the width of the fascial windows (aponeuroses) in live subjects using real-time ultrasound (US) imaging. On the basis of de Se ze and Cazelets’ findings (de Se ze and Cazalets, 2008), the aponeuroses at T3 and L4 were selected as potential sites for sEMG of the thoracic and lumbar ES, respectively. Measurements were made on a group of younger ( o30 years) and a group of older ( 470 years) participants to assess the effect of aging.
2. Methods A power analysis based on de Se ze and Cazelets0 results indicated that ten subjects in each age group would be sufficient to address our hypotheses. Accordingly, 20 healthy volunteers were recruited for this study by advertisement in a hospital. Ten subjects were aged between 18 and 30 years and ten were over 70 years, with equal numbers of males and females in each group (Table 2). Exclusion criteria included previous significant spinal problems such as fracture, surgery or disease; thoracotomy; pregnancy, neuropathy or myopathy. All subjects gave informed consent. Ethics approval for this study was granted by the ACT Health Human Research Ethics Committee and the Australian National University Human Research Ethics Committee. Ultrasonography was performed by DP or JS under the guidance of a trained sonographer using a Mindray DP-6600 ultrasound scanner with a 3.5 MHz R50 electronic convex array transducer (Mindray Biomedical, Shenzhen, China). The participants were positioned in prone lying. After palpating all of the spinous processes, the T3 and L4 spinous processes were marked with ink. The US transducer head was always positioned to the right of the T3 and L4 spinous processes because aponeurosis widths have previously been reported to be symmetrical (de Se ze and Cazalets, 2008). US images were acquired in each of
D.M. Perriman et al. / Journal of Biomechanics 44 (2011) 1025–1030
Table 2 The subject demographics, mean 7 standard deviation and [range]. The younger (o 30) and the older (470) groups each included 5 males and 5 females. Age category
Age (years)
Weight (kg)
Height (cm)
o30 470
23.9 7 1.8 [21–27] 76.1 7 4.6 [70–84]
68.3 7 10.8 [55–91] 73.2 7 8.7 [57–88]
173.6 711.1 [157–190] 173.3 711.5 [161–197]
Fig. 1. US images of the T3 spinous process (SP) with the attachments of the right trapezius (Tr), rhomboids (Rh) and the erector spinae (ES): (A) at rest and (B) in extension. Arrows indicate the points between which the aponeurosis width measurement was made. Three layers of fascia are identified: the trapezius aponeurosis (a), the rhomboid aponeurosis which is continuous with the superficial lamina of the posterior layer of the thoraco-lumbar fascia (b), and the deep lamina of the posterior layer of the thoraco-lumbar fascia (c). The trapezius aponeurosis was seen to be separate from the rhomboid aponeurosis during extension. three positions: (i) arms above the head (‘above head’), (ii) arms hanging by the side (‘hanging’) and (iii) arms lying alongside the body (‘by side’). These positions were chosen in order to examine the effect of changes in superficial muscle position on the T3 aponeurosis width. The order in which the subjects assumed
1027
each position was randomized and the locations of the spinous processes were reassessed after each change of position. US images were acquired at rest and then again during sustained back extension from prone for each of the three positions, resulting in a total of six thoracic images for each subject. Six lumbar images were then recorded in the same way. Measurements of aponeurosis width were made offline from the US images. At T3, aponeurosis width measurements were made by identifying the hyperechoic fascial lines situated above and below the trapezius muscle and measuring the distance between the point where they coalesced (the medial edge of the muscle) and the tip of the spinous process (Fig. 1A and B). L4 aponeurosis width measurements were made by identifying the fascial lines above and below the latissimus dorsi and measuring the distance between the point just lateral to where they coalesced and the tip of the spinous process (Fig. 2A and B). If an image was insufficiently clear to delineate the precise point at which the fascia coalesced, a conservative judgment in favor of a greater aponeurosis width was
Fig. 2. US images of the L4 spinous process (SP) with the attachments of the right latisimus dorsi (LD) and the erector spinae (ES). (A) at rest and (B) in extension. Arrows indicate the points between which the aponeurosis width measurement was made. Two layers of fascia are identified: the latisimus dorsi aponeurosis which is continuous with the superficial lamina of the posterior layer of the thoraco-lumbar fascia (a), and the deep lamina of the posterior layer of the thoraco-lumbar fascia (b). The two layers were pushed upwards during extension into a domed shape due to the contraction of the ES resulting in a narrowing of the aponeurotic window.
1028
D.M. Perriman et al. / Journal of Biomechanics 44 (2011) 1025–1030
Table 3 Measurements of aponeurosis width (mm) for the whole group, those o 30 years of age (n¼ 10) and 470 years (n ¼10) at T3 and L4. The measurements at rest and during extension in prone are presented separately.
T3 Rest Mean 7 SD Range Extended Mean 7 SD Range L4 Rest Mean 7 SD Range Extended Mean 7 SD Range
All subjects
o 30 years
Females and males
Female
Male
Female
Male
4.4 7 4.7 0–21.2
5.6 75.8 0–21.2
0.37 0.7 0–2.3
5.9 73.8 0–10.7
5.6 7 4.6 0–17.4
1.8 7 2.6 0–9.1
2.3 72.3 0–8.2
0.27 0.4 0–1.2
3.4 73.3 0–9.1
1.4 7 2.2 0–5.9
35.5 7 7.0 18–52.8
36.8 7 4.6 28.1–44.9
40.47 4.8 29.8–48.7
30.3 75.2 18–38.8
34.7 7 8.9 25.2–52.8
29.4 7 7.1 14–52.6
25.9 7 8.4 19.6–35.3
33.17 5.4 24.9–42.9
25.5 7 6.1 14–33.2
31.6 7 9.5 22.8–52.6
470 years
made. All measurements were made using the linear measurement software incorporated within the Mindray system (Figs. 1 and 2).
>70
<30 50
The data for each group were expressed as mean and standard deviations (mean (7 SD)). A split-plot ANOVA model with age group and gender as between-subject effects, and activity (at rest and in extension), arm position and arm position order as within-subject effects, was used for both thoracic and lumbar data. Differences were considered significant when p o0.05.
aponeurosis width (mm)
2.1. Statistical analysis
40 30 20 10
3. Results 3.1. Thoracic (T3) The mean aponeurosis width between the tip of the spinous process of T3 and the adjacent trapezius muscle was significantly smaller during extension than at rest (1.872.6 mm vs. 4.474.7 mm, p o0.0001, Table 3). Neither arm position nor the order in which the positions were assumed had a significant effect. Males had significantly smaller mean aponeurosis widths than females (p¼0.049, Table 3 and Fig. 3). This difference was most marked in young males who had measurements approaching zero during both rest and extension, contrasting with young females and both males and females in the older group, all of whom had mean aponeurosis widths of between 5.6 and 5.9 mm at rest (Table 3 and Fig. 3). 3.2. Lumbar (L4) The mean aponeurosis width between the tip of the spinous process of L4 and the medial edge of the adjacent latissimus dorsi muscle was significantly smaller during extension than at rest (29.477.1 mm vs. 35.5 77.0 mm, po0.0001, Fig. 3, Table 3). The order in which the positions were assumed did not have a significant effect. 3.3. The dynamic behavior of the muscle with respect to the aponeurosis The appearance and behavior of the aponeuroses in the thoracic and lumbar regions were very different. The most superficial muscle at the thoracic site was the trapezius. At rest, this muscle was attached to the T3 spinous process via a clearly visible
0 female
male
female
Thoracic Rest
Thoracic Extension
Lumbar Rest
Lumbar Extension
male
Fig. 3. Boxplot of the aponeurosis widths at T3 and L4 during rest and extension for the males and females in both the younger and older age groups. The aponeurosis width at rest was significantly wider than in extension at both sites (p o 0.0001). Males had significantly narrower T3 aponeurosis widths than females (p o 0.05).
aponeurosis in 15 out of the 20 subjects. In the remaining five subjects the muscle inserted directly onto the spinous process with no discernable aponeurosis. Where an aponeurosis was present, its width decreased during active extension. In three of the 15 subjects who had a discernable aponeurosis at rest, it was reduced to nothing during extension. When present, the aponeurosis appeared to connect both the trapezius and the rhomboid muscles via a single structure at rest, but during extension it was seen to cleave horizontally as the trapezius muscle fibers thickened within the space revealing separate attachments. In the five subjects whose trapezius appeared to insert directly onto the spinous process, this thickening was profound. The rhomboid muscle, which was positioned deep to the trapezius, increased in thickness but always inserted onto the spinous process via an aponeurosis situated below the trapezius muscle or the trapezius aponeurosis (Fig. 1B). In contrast to the T3 aponeurosis, the L4 aponeurosis was wide and thick and did not appear to cleave during extension but was seen to bulge or dome posteriorly over the thickened ES (Fig. 2B). The latissimus dorsi muscle belly was seen to thicken, but its relation to the aponeurosis did not differ markedly from that seen in the resting position (Fig. 2A and B).
D.M. Perriman et al. / Journal of Biomechanics 44 (2011) 1025–1030
4. Discussion The aim of this study was to investigate the utility of two recently proposed sEMG recording sites for the ES in vivo. There were four major findings. First, the most superficial aponeurosis width adjacent to T3 was either non-existent or too narrow to allow sEMG recording. Second, aponeurosis widths at both T3 and L4 reduced significantly during active back extension. Third, aponeurosis width at T3 was significantly narrower in men than in women with young men often demonstrating no discernable aponeurosis at all. Finally, the trapezius muscle attached to the spinous process at T3 within an envelope of connective tissue which was separate to the aponeurosis of the rhomboid muscle lying below it. These findings have not previously been reported in the literature. The accuracy of direct recording of muscle sEMG amplitudes depends critically on the positioning of the recording electrodes so as to avoid crosstalk from other muscles (Soderberg and Knutson, 2000). The ES have been studied using sEMG for nearly 60 years (Table 1). Surface EMG recording electrodes for the thoracic ES are most commonly sited in the lower thoracic spine because of concerns over crosstalk from the trapezius and rhomboid muscles (Floyd and Silver, 1951). The most common site used is 40–50 mm lateral to T9 (McGill, 1991; Callaghan and Dunk, 2002). This site was first suggested by Lafortune et al. (1988) who found that sEMG signal amplitudes measured here and 30 mm adjacent to L4 were the same. These recordings were made from eight young men while they performed squat lifts (Lafortune et al., 1988). Until now this site has been the accepted standard for electrode placement. As noted previously, recent anatomical evidence has suggested that recordings made adjacent to T9 are likely to be subject to significant crosstalk (de Se ze and Cazalets, 2008). Cadaveric dissection revealed that the thoracic ES were covered by superficial muscle tissue from T5 to T11 but de Se ze and Cazalets suggested that there was an ‘electrically silent window’ between C7 and T4 through which the thoracic ES could potentially be recorded. This window was widest at T3 with a mean width of 34 mm. They recommended positioning sEMG electrodes 20 mm from the midline at T3. Other investigators have located their electrodes 25 mm adjacent to the T3 spinous process (Seelen et al., 1998). Our in vivo study demonstrated, however, that the mean T3 aponeurosis width at rest was only 6 mm reducing to 3 mm during active back extension. Thus, according to our results the sites suggested by de Se ze and Cazalets and those used by previous researchers would lie over the belly of the trapezius muscle. The T3 site is therefore not suitable for sEMG recording of the thoracic ES in vivo, especially during functional activities, unless the trapezius and rhomboid muscles are specifically rendered completely inactive. In the light of both our results and those of de Se ze and Cazalets the findings of previous studies which have used sEMG to record the thoracic ES may need reconsideration (Table 1). Although difficult to obtain directly, the activity of deeply situated muscles can be inferred from sEMG. McGill et al. (1996) hypothesized that deep trunk muscle activity could be predicted from surface activity because the neural drive to both is matched. They found that the difference in amplitude between superficial and deep recordings was generally within 15% and argued that this magnitude of error was within acceptable limits for biomechanical modeling (McGill et al., 1996). Whether there is a uniform relationship between the surface muscles and the deeper thoracic ES has not yet been established. The suitability of L4 as a recording site is not in dispute and our findings confirm that the aponeurosis adjacent to L4 is sufficiently wide for direct sEMG recording. de Se ze and Cazalets
1029
recorded a mean aponeurosis width of 86 mm and recommended that sEMG electrodes be sited at a distance of 40 mm from the midline at L4 (de Se ze and Cazalets, 2008). Many previous studies have used distances of between 20 and 30 mm (Table 1). Importantly, our study showed that the mean aponeurosis width reduced significantly in extension from 35.5 mm to just 29 mm. This reduction in aponeurosis width was the result of a ‘doming’ effect produced by the contraction of the underlying ES during active extension (Fig. 2B). In fact, this ‘doming’ can easily be observed from the skin surface but this is the first time that the effect on the aponeurotic window has been reported. Our results indicate that if subjects are to be studied during extension a distance of 20 mm from the L4 spinous process would ensure that the electrodes were positioned over the aponeurosis, thereby preventing crosstalk from the latissimus dorsi. The finding that the lumbar aponeurosis width may be reduced during strong contraction of the erector spinae is relevant to the findings of Lafortune et al. (1988). As previously noted, their assertion that the T9 site records the thoracic ES rests on the observation that lumbar ES signal amplitudes measured at 30 mm adjacent to L4 were the same as signal amplitudes measured 40 mm lateral to T9. In our study the mean aponeurosis width at L4 in young males during extension was 33 mm so siting the recording electrode at 30 mm may have resulted in significant crosstalk from the latissimus dorsi in their study. Further, the site adjacent to T9 is completely covered by the muscle fibers of either the trapezius or latissimus dorsi (de Se ze and Cazalets, 2008) raising the strong possibility that the signal detected by Lafortune et al. may actually have arisen from the latissimus dorsi which would have been strongly activated during the squat-lift protocol used in their study. The US images of the aponeurosis showed that the rhomboid and trapezius muscles were surrounded by separate fascial layers which, in most cases, coalesced to form their aponeurotic attachments onto the spinous process. For trapezius the extension task demonstrated this arrangement very clearly: the muscle expanded within its fascial envelope during contraction, dramatically reducing the fascial window and removing the possibility of any ‘electrically silent’ access to the ES muscles below. This configuration is consistent with descriptions of the trapezius lying between two laminae of the deep cervical fascia (Newell, 2008). The rhomboid aponeurosis did not appear to reduce in length during the extension task. The images of the lumbar aponeurosis were consistent with previous anatomical descriptions (Bogduk and Macintosh, 1984; Vleeming et al., 1996; Barker and Briggs, 1999; Loukas et al., 2008). The limitations of this study include small sample size and limited US quality. Though small, the number of subjects used in this study was sufficient to address its main aim. A larger cohort might further clarify the age and gender effects which we observed. A 3.5 MHz transducer was used to acquire the images for this study. A higher frequency US unit would undoubtedly have resulted in more detailed images (Salmhofer et al., 1996). To overcome this potential limitation, we were conservative in our measurements in cases where clarity of image was an issue, reporting larger width measurements in all cases. We are therefore confident that the results do not underestimate aponeurosis width. Finally, because the study participants were only positioned in prone lying extrapolating our findings to other positions should be done with caution.
5. Conclusion This study reveals the potential risks of extrapolating findings from cadaveric studies to living systems. It transpires that T3 is
1030
D.M. Perriman et al. / Journal of Biomechanics 44 (2011) 1025–1030
not a reliable site over which to record the ES directly in vivo because the trapezius muscle, when contracted, covers the ES. This is consistent with the suggestion that many previous studies which have purported to measure the thoracic ES using sEMG have, in fact, recorded activity from the superficial muscles. In the lumbar spine we would recommend electrode placement no further than 20 mm adjacent to L4. On the basis of our results, future researchers are encouraged to re-examine the use of sEMG to study the ES. In theory, nEMG would be more accurate but the difficulties associated with its use make it less practical. An alternative approach would be to use sEMG and compensate for crosstalk from superficial muscles by employing techniques which use a larger number of recording sites along with more advanced signal processing analysis.
Conflict of interest statement None of the authors have any financial and personal relationships with other people or organizations that could have inappropriately influenced this work including employment, consultancies, stock ownership, honoraria, paid expert testimony and patent applications/registrations. Funding for this work was provided by an NH&MRC PhD scholarship, the Canberra Hospital Private Practice Fund and AO Spine Pacific.
Acknowledgments The authors would like to thank the study participants, Roxanne Miller and Annette Lawson for their invaluable assistance with this study, and also Dr. Nick Ball for his very helpful advice. We gratefully acknowledge the support of the NHMRC Australia, AO Spine Australasia and the Canberra Hospital Private Practice Fund. We would also like to respectfully acknowledge the anonymous reviewers who greatly enhanced this manuscript with their thoughtful suggestions. References Abdoli-E, M., Agnew, M.J., et al., 2006. An on-body personal lift augmentation device (PLAD) reduces EMG amplitude of erector spinae during lifting tasks. Clinical Biomechanics 21 (5), 456–465. Barker, P.J., Briggs, C.A., 1999. Attachments of the posterior layer of lumbar fascia. Spine 24 (17), 1757–1764. Bogduk, N., Macintosh, J.E., 1984. The applied anatomy of the thoracolumbar fascia. Spine 9 (2), 164–170. Callaghan, J.P., Dunk, N.M., 2002. Examination of the flexion relaxation phenomenon in erector spinae muscles during short duration slumped sitting. Clinical Biomechanics (Bristol, Avon) 17 (5), 353–360. Cholewicki, J., Lee, A.S., et al., 2009. Trunk muscle response to various protocols of lumbar traction. Manual Therapy 14 (5), 562–566. Cholewicki, J., McGill, S.M., 1996. Mechanical stability of the in vivo lumbar spine: implications for injury and chronic low back pain. Clinical Biomechanics (Bristol, Avon) 11 (1), 1–15. de Se ze, M.P., Cazalets, J.R., 2008. Anatomical optimization of skin electrode placement to record electromyographic activity of erector spinae muscles. Surgical and Radiolic Anatomy 30 (2), 137–143.
Floyd, W.F., Silver, P.H., 1951. Function of erectores spinae in flexion of the trunk. Lancet 1 (6647), 133–134. Floyd, W.F., Silver, P.H., 1955. The function of the erectores spinae muscles in certain movements and postures in man. Journal of Physiology 129 (1), 184–203. Fuglevand, A.J., Winter, D.A., et al., 1992. Detection of motor unit action potentials with surface electrodes: influence of electrode size and spacing. Biological Cybernetics 67 (2), 143–153. Greig, A.M., Bennell, K.L., et al., 2008. Postural taping decreases thoracic kyphosis but does not influence trunk muscle electromyographic activity or balance in women with osteoporosis. Manual Therapy 13 (3), 249–257. Keenan, K.G., Farina, D., et al., 2007. Sensitivity of the cross-correlation between simulated surface EMGs for two muscles to detect motor unit synchronization. Journal of Applied Physiology 102 (3), 1193–1201. Krajcarski, S.R., Potvin, J.R., et al., 1999. The in vivo dynamic response of the spine to perturbations causing rapid flexion: effects of pre-load and step input magnitude. Clinical Biomechanics (Bristol, Avon) 14 (1), 54–62. Lafortune, D., Norman, R., et al., 1988. Ensemble averages linear enveloped EMGs during lifting. In: Proceedings of the Biannual Conference of the Canadian Society for Biomechanics, Ottawa, Canada. Loukas, M., Shoja, M.M., et al., 2008. Anatomy and biomechanics of the vertebral aponeurosis part of the posterior layer of the thoracolumbar fascia. Surgical and Radiolic Anatomy 30 (2), 125–129. Martin, D.C., Medri, M.K., et al., 2001. Comparing human skeletal muscle architectural parameters of cadavers with in vivo ultrasonographic measurements. Journal of Anatomy 199 (Part 4), 429–434. McGill, S., Juker, D., et al., 1996. Appropriately placed surface EMG electrodes reflect deep muscle activity (psoas, quadratus lumborum, abdominal wall) in the lumbar spine. Journal of Biomechanics 29 (11), 1503–1507. McGill, S.M., 1991. Electromyographic activity of the abdominal and low back musculature during the generation of isometric and dynamic axial trunk torque: implications for lumbar mechanics. Journal of Orthopedic Research 9 (1), 91–103. McGill, S.M., 1992. The influence of lordosis on axial trunk torque and trunk muscle myoelectric activity. Spine 17 (10), 1187–1193. Meekins, G.D., So, Y., et al., 2008. American Association of Neuromuscular and Electrodiagnostic Medicine evidenced-based review: use of surface electromyography in the diagnosis and study of neuromuscular disorders. Muscle and Nerve 38 (4), 1219–1224. Moseley, G.L., Hodges, P.W., et al., 2003. External perturbation of the trunk in standing humans differentially activates components of the medial back muscles. Journal of Physiology 547 (Part 2), 581–587. Newell, L., 2008. Section 5—the back. In: Standring, S. (Ed.), Gray’s Anatomy 40th Edn. Churchill Livingstone, New York, pp. 705–773. O0 Sullivan, P.B., Dankaerts, W., et al., 2006. Effect of different upright sitting postures on spinal-pelvic curvature and trunk muscle activation in a pain-free population. Spine 31 (19), E707–E712. Salmhofer, W., Rieger, E., et al., 1996. Influence of skin tension and formalin fixation on sonographic measurement of tumor thickness. Journal of the American Academy of Dermatology 34 (1), 34–39. Schultz, A., Andersson, G.B., et al., 1982. Analysis and quantitative myoelectric measurements of loads on the lumbar spine when holding weights in standing postures. Spine 7 (4), 390–397. Seelen, H.A., Potten, Y.J., et al., 1998. Development of new muscle synergies in postural control in spinal cord injured subjects. Journal of Electromyography and Kinesiology 8 (1), 23–34. Soderberg, G.L., Knutson, L.M., 2000. A guide for use and interpretation of kinesiologic electromyographic data. Physical Therapy 80 (5), 485–498. Tsui, B., Dillane, D., et al., 2008. Ultrasound imaging in cadavers: training in imaging for regional blockade at the trunk. Canadian Journal of Anesthesia 55 (2), 105–111. Vleeming, A., Pool-Goudzwaard, A.L., et al., 1996. The function of the long dorsal sacroiliac ligament: its implication for understanding low back pain. Spine 21 (5), 556–562. Vleeming, A., Pool-Goudzwaard, A.L., et al., 1995. The posterior layer of the thoracolumbar fascia. Its function in load transfer from spine to legs. Spine 20 (7), 753–758. Young, C.C., Rose, S.E., et al., 1989. The effect of surface and internal electrodes on the gait of children with cerebral palsy, spastic diplegic type. Journal of Orthopedic Research 7 (5), 732–737.