Do surface electrode recordings validly represent latissimus dorsi activation patterns during shoulder tasks?

https://doi.org/10.1016/j.jelekin.2014.10.008Get rights and content

Abstract

Because of its superficial location surface electrodes are commonly used to record latissimus dorsi (LD) activity. Despite the fact that the recommended electrode placement is over the belly where LD is quite thin no studies have investigated the possibility of signal contamination from muscles lying deep to LD. Therefore, the aim of this study was to determine the validity of using surface electrodes to record activity from LD. Eight asymptomatic subjects performed ramped isometric (0–100% maximum load) and dynamic (70% maximum load) shoulder tasks. Intramuscular electrodes were inserted into LD and the adjacent erector spinae. Surface electrodes were placed over LD around the intramuscular electrodes. Results indicated that while there was no difference in activity level or activation pattern (ICC > 0.94) recorded by the two electrode types during shoulder tasks in which LD would be expected to be active (extension and adduction), significantly lower (p < 0.05) LD activity was recorded via intramuscular electrodes during the shoulder flexion and abduction tasks. Therefore, recordings of LD activity by surface electrodes overestimate LD activity during shoulder tasks when this muscle would be expected to be activated at minimum levels. Erector spinae immediately deep to LD was confirmed as a source of crosstalk contamination.

Introduction

Latissimus dorsi (LD) is a large, flat, triangular muscle extending from a broad attachment on the pelvis and thoracolumbar vertebrae to a tendinous attachment into the bicipital groove on the humerus. It is a major extensor, adductor and internal rotator of the shoulder joint. Because of its superficial placement on the back surface electrodes are commonly used to record LD activity. A simple, electronic search of the literature between 2011 and 2013 found over 20 papers reporting LD activity recorded with the use of surface electrodes. The scope of these studies ranged from investigating LD activity during: normal shoulder function (Hawkes et al., 2012b, Rota et al., 2013); various sporting activities including rowing (Bazzucchi et al., 2013), tennis (Rota et al., 2012), pole vaulting (Frere et al., 2012), skiing (Nilsson et al., 2013), baseball (Reyes et al., 2011), golf (Lim et al., 2012) and swimming (Ikuta et al., 2012); exercises (Marchetti and Uchida, 2011); shoulder and trunk dysfunction (Hawkes et al., 2012a, Ntousis et al., 2013, Wood et al., 2011); classical singing (Watson et al., 2012); and muscle fatigue (Balasubramanian et al., 2014).

Muscle activity can be recorded by surface electrodes that attach to the skin over the muscle of interest or by indwelling (intramuscular) electrodes that are inserted into the muscle. Surface electrodes have the advantage of being non-invasive and sample from a larger cross-section of the muscle of interest than indwelling electrodes. However, their larger pick-up volume means that surface electrodes may detect signals from neighboring muscles lying adjacent or deep to the muscle of interest. This phenomenon, known as crosstalk, is the most significant limiting factor in using surface electrodes to record muscle activity.

In order to reduce the risk of crosstalk from adjacent muscles surface electrodes are commonly placed at the widest section of the muscle of interest. In the case of LD this is on the back over the center of the muscle belly 4 cm below the inferior angle of the scapula (Cram et al., 1998, Nilsson et al., 2013, Ntousis et al., 2013, Park and Yoo, 2013). However, at this position LD is much thinner than near its narrow superior section where it forms part of the posterior axillary wall (Williams et al., 1989). This common electrode placement position therefore, makes surface electrode recordings from LD potentially susceptible to crosstalk from muscles lying deep to LD e.g. erector spinae (ES). Despite this obvious, potential source of signal contamination no studies have assessed the validity of using surface electrodes placed over the muscle belly 4 cm below the inferior angle to record activity from LD. As our quick search of the literature indicates there is significant interest in examining activity patterns in LD in a variety of fields. There is thus an urgent need to confirm that surface electrode recordings from latissimus dorsi are not contaminated by crosstalk and truly represent activity from this muscle.

Therefore, the aim of this study was to determine the validity of using surface electrodes to record activity from LD during isometric and dynamic tasks in which LD would be expected to be highly activate (shoulder extension and adduction) and those in which it would be expected to be activated at low levels (shoulder flexion and abduction). If evidence of crosstalk contamination was found in the surface electrode recording from LD, a secondary aim was to determine if ES was the source of this contamination.

Section snippets

Participants

Eight asymptomatic participants (five male, three female, aged 19–49 years) who had had no pain in their dominant shoulder over the previous two years nor had ever been treated for shoulder pain volunteered to participate in this investigation. A power calculation using G∗Power (Faul et al., 2007) showed that eight subjects at an α = 0.05, β = 0.80 would allow the detection of an effect size of 1.16 or, in other words, a difference in the mean EMG signals between the electrode types of approximately

Results

Mean (±95% CI) LD activity levels recorded from both surface and intramuscular electrodes and ES activity levels for each of the test conditions examined are depicted in Fig. 2. During the different ramped isometric tests, there was a significant main effect between LD activity recorded from different tests (F3,21 = 3.40, p = 0.04), between different electrode types (F1,7 = 9.8, p = 0.02) and an interaction between test and electrode type (F3,21 = 5.46, p = 0.006). Tukey post hoc analysis indicated that LD

Discussion

The results of this study indicate that activity levels in LD can vary considerably depending on whether they are recorded using intramuscular or surface electrodes. While there was no difference in LD activity levels recorded by the two electrode types during any of the tasks in which LD would be expected to be active (shoulder extension and adduction), significantly lower (p < 0.05) LD activity was recorded via intramuscular electrodes than that recorded via surface electrodes during both the

Conflict of interest

The authors declare that they have no conflict of interest.

Karen Ginn is an Associate Professor in the Discipline of Biomedical Science in the Sydney Medical School at the University of Sydney and a musculoskeletal physiotherapist in part-time private clinical practice. Her research interest is the shoulder: EMG studies, clinical trials, experimental pain studies and investigations of shoulder stiffness and cortical changes associated with shoulder pain.

References (29)

  • I. Bazzucchi et al.

    Cardio-respiratory and electromyographic responses to ergometer and on-water rowing in elite rowers

    Eur J App Physiol

    (2013)
  • J.S. Bendat et al.

    Random data: analysis and measurement procedures

    (1971)
  • C. Boettcher et al.

    Standard maximum isometric voluntary contraction tests for normalizing shoulder muscle EMG

    J Orthop Res

    (2008)
  • J. Cram et al.

    Introduction to surface electromyography

    (1998)
  • Cited by (24)

    • Activation patterns of shoulder internal and external rotators during pure axial moment generation across a postural range

      2021, Journal of Biomechanics
      Citation Excerpt :

      An important limitation of this experiment is the lack of intramuscular EMG for several key shoulder muscles. Indwelling electrodes are being increasingly recommended for recording of activity for many shoulder muscles (Ginn & Halaki, 2015; Johnson et al., 2011), as surface electrodes have been shown to overestimate activity in some muscles, such as the infraspinatus and latissimus dorsi, during tasks where they would be expected to be minimally active. This may have implications for the interpretation of the internal rotation results from the current study.

    • Do surface electrodes validly represent lower trapezius activation patterns during shoulder tasks?

      2020, Journal of Electromyography and Kinesiology
      Citation Excerpt :

      The scope of these papers ranged from investigating lower trapezius activity in: shoulder exercises (Arshadi et al., 2019; Grant et al., 2019; Ho et al., 2019; Joseph et al., 2019; Kim and Yoo, 2019; Park and Park, 2019; Pozzi et al., 2019; Scott et al., 2018), shoulder movements (Contemori et al., 2019; Mendez-Rebolledo et al., 2019), tennis players (Gillet et al., 2019), gymnasts (Kochanowicz et al., 2019), patients with shoulder pain (Kara et al., 2019; Kim et al., 2019; Seitz et al., 2019), swimmers (Sabzehparvar et al., 2019) and table tennis players (Meghdadi et al., 2019) with shoulder pain, and paraplegic gait (Baniasad et al., 2018). A recent study investigating the validity of surface electromyography to record activity from latissimus dorsi, another superficial shoulder muscle on the upper back, found significant crosstalk contamination from the underlying erector spinae muscle group (Ginn and Halaki, 2015). These authors concluded that, as the activity recorded by surface electrodes did not accurately represent latissimus dorsi activity in some shoulder tasks, it is not valid to use this electrode type to investigate latissimus dorsi activation patterns during complicated functional shoulder tasks.

    • Normalising surface EMG of ten upper-extremity muscles in handcycling: Manual resistance vs. sport-specific MVICs

      2020, Journal of Electromyography and Kinesiology
      Citation Excerpt :

      Hence, sport-specific MVIC normalisation of LD amplitude is a suitable approach in handcycling. However, it must be noted that LD activation using sEMG electrodes is overestimated when compared to intramuscular electrodes due to crosstalk contamination (Ginn and Halaki, 2015). For RA, lower amplitudes were found during sport-specific trials.

    • Static optimization underestimates antagonist muscle activity at the glenohumeral joint: A musculoskeletal modeling study

      2019, Journal of Biomechanics
      Citation Excerpt :

      Interelectrode distances were 20 mm, and skin impedance verified to be less than 5 kΩ (Day, 2002). Bipolar intramuscular (in-dwelling) electrodes were placed in anterior, middle and posterior deltoid sub-regions, rhomboid major, supraspinatus, infraspinatus, subscapularis, pectoralis minor, serratus anterior, teres major, and latissimus dorsi (Boettcher et al., 2008; Johnson et al., 2011; Ginn and Halaki, 2015), with ultrasonic guidance employed for rhomboid major and pectoralis minor electrode placement (Mindray, DP-9900). A ground electrode was placed on the contralateral acromion and scapular spine (Universal Electrosurgical Pad: Split, 9160F, Australia).

    View all citing articles on Scopus

    Karen Ginn is an Associate Professor in the Discipline of Biomedical Science in the Sydney Medical School at the University of Sydney and a musculoskeletal physiotherapist in part-time private clinical practice. Her research interest is the shoulder: EMG studies, clinical trials, experimental pain studies and investigations of shoulder stiffness and cortical changes associated with shoulder pain.

    Mark Halaki is currently a senior lecturer in the Discipline of Exercise and Sport Science, The Faculty of Health Sciences, The University of Sydney. He has a background in mechanical and biomedical engineering with research interests in electromyographic studies in the areas of motor control and biomechanics.

    View full text