Measurement of scapular dyskinesis using wireless inertial and magnetic sensors: Importance of scapula calibration

Journal of Biomechanics ∎ (∎∎∎∎) ∎∎∎–∎∎∎

Contents lists available at ScienceDirect

Journal of Biomechanics journal homepage: www.elsevier.com/locate/jbiomech www.JBiomech.com

Measurement of scapular dyskinesis using wireless inertial and magnetic sensors: Importance of scapula calibration Josien C. van den Noort a,b,n, Suzanne H. Wiertsema a, Karin M.C. Hekman a, Casper P. Schönhuth c, Joost Dekker a,d, Jaap Harlaar a,b a

VU University Medical Center, Department of Rehabilitation Medicine, PO Box 7057, 1007 MB Amsterdam, The Netherlands MOVE Research Institute Amsterdam, The Netherlands c VU University Medical Center, Department of Orthopaedics, Amsterdam, The Netherlands d EMGO Institute for Health and Care Research, Amsterdam, The Netherlands b

art ic l e i nf o

a b s t r a c t

Article history: Accepted 24 May 2015

Measurement of 3D scapular kinematics is meaningful in patients with shoulder pathologies showing scapular dyskinesis. This study evaluates the effect of single and double anatomical calibration (0° and 120°) with a scapula locator compared to standard calibration (using sensor alignment with the spina scapulae and static upright posture, ISEO-protocol) on 3D scapular kinematics measured with an inertial and magnetic measurement system (IMMS). Ten patients with scapular dyskinesis performed humeral anteflexion and abduction movements while 3D scapular kinematics were measured using IMMS sensors. The sensor on the scapula was anatomically calibrated (i) according to the ISEO-protocol, (ii) using single scapula locator calibration (0°) and (iii) double scapula locator calibration (0° and 120°). For calibration, the scapula locator (with IMMS) was positioned on the scapula, while holding the humerus at several anteflexion and abduction postures. Single and double calibration resulted in a significant increase of scapular anterior tilt (14–30°) with respect to the skin-fixed sensor (ISEO). Protraction angles were not significantly different. During anteflexion, double calibration did not show a significant increase in lateral rotation compared to single calibration. During abduction of 4 90°, double calibration showed 10–14° increased lateral rotation with respect to single calibration, although this was not significant (P40.06). Calibration with a scapula locator when applying IMMS is necessary, because measures of scapular anterior tilt are grossly underestimated with the ISEO-protocol. For shoulder movements that exceed 90° elevation, a double calibration prevents small but relevant underestimation of lateral rotation angles of the scapula. & 2015 Elsevier Ltd. All rights reserved.

Keywords: Scapula Shoulder Movement analysis Scapular dyskinesis Inertial and magnetic sensors

1. Introduction Measurement of 3D scapular kinematics is particularly meaningful in patients with shoulder pathologies showing scapular dyskinesis (Kibler et al., 2009; Kibler and Sciascia, 2010; Ludewig and Reynolds, 2009; McClure et al., 2009; Tate et al., 2009; van den Noort et al., 2014). Scapular dyskinesis is defined as a posterior displacement of the scapular medial border and/or inferior angle away from the thorax (winging) or a dysrhythmia of the scapular motion, such as premature or excessive elevation or protraction during arm elevation, a rapid medial rotation during arm lowering, or a non-smooth motion during arm elevation or lowering n Corresponding author at: VU University Medical Center, Department of Rehabilitation Medicine, PO Box 7057, 1007 MB Amsterdam, The Netherlands. Tel.: þ 31 20 444 3192; fax: þ31 20 444 0787. E-mail address: [email protected] (J.C. van den Noort).

(McClure et al., 2009). Such alterations have been observed in e.g. shoulder instability, rotator cuff injury and impingement syndrome (De Baets et al., 2013; Lukasiewicz et al., 1999; McClure et al., 2006; Roren et al., 2013; Struyf et al., 2011; Warner et al., 1992). Wireless sensors of an inertial and magnetic measurement system (IMMS) are suitable to conveniently measure the 3D kinematics of the scapula. A few studies evaluated the intra- and inter-observer reliability and precision of such a system in healthy subjects (Cutti et al., 2008; Parel et al., 2014, 2012; van den Noort et al., 2014). Technical dynamic accuracy of IMMS sensors used in the latter study is reported to be around 2° (Xsens Technologies B. V., 2011). Standard errors of measurement (SEM) of IMMS in scapular kinematic measurement were found to be within 5° for both intra-and inter-observer data of medio/lateral rotation, anterior/posterior tilt and for intra-observer data of scapular re/ protraction (van den Noort et al., 2014). Inter-observer data of re/

http://dx.doi.org/10.1016/j.jbiomech.2015.05.036 0021-9290/& 2015 Elsevier Ltd. All rights reserved.

Please cite this article as: van den Noort, J.C., et al., Measurement of scapular dyskinesis using wireless inertial and magnetic sensors: Importance of scapula calibration. Journal of Biomechanics (2015), http://dx.doi.org/10.1016/j.jbiomech.2015.05.036i

J.C. van den Noort et al. / Journal of Biomechanics ∎ (∎∎∎∎) ∎∎∎–∎∎∎

2

protraction showed a SEM of 6–8°. Palpation errors of bony landmarks of the shoulder are shown to be about 2° (de Groot, 1997) whereas reliability of a scapula locator has been shown to be within 4° (Barnett et al., 1999). Using IMMS, anatomical calibration of the sensor placed on the skin with respect to the underlying scapular bone (sensor-tosegment orientation) can be based on alignment of the sensor with the spina scapulae and the gravity vector, such as described in (Cutti et al., 2008) and applied by (van den Noort et al., 2014) (ISEO-protocol). However, according to the ISEO-protocol, anatomical alignment of the scapula sensor is crucial as well as the measure of a static upright posture. Therefore, differences between examiners may result in inaccuracies. This was in particular seen in offsets and change in range of motion in anterior/ posterior tilt and re/protraction (van den Noort et al., 2014). Furthermore, soft tissue artifacts of the loose skin, fat tissue and bulging deltoid muscles might cause a difference between sensor movement and actual scapula movement in all 3 planes of movement, especially for higher humeral elevation angles (Brochard et al., 2011; Meskers et al., 2007; Shaheen et al., 2011b; van Andel et al., 2009). To improve the anatomical calibration, a scapula locator with adjustable bars (a tripod, placed on the angulus inferior, angulus acromialis and trigonum scapulae), including a sensor, could be used (Johnson et al., 1993). A scapula locator has been formerly used to validate an acromion marker cluster (van Andel et al., 2009) and an electromagnetic tracking device on the acromion (Meskers et al., 2007) that aimed to track scapular motion. Meskers et al. (2007) suggested that the locator should be used to calibrate such skin-fixed sensors. A similar approach, i.e. the use of an external frame on palpable anatomical landmarks for calibration, has been previously described for IMMS in the measurement of lower limb kinematics (Picerno et al., 2008). The calibration position of the scapula locator, i.e. at which degree of humeral elevation, is shown to be important for accurate measurement of scapular kinematics (Prinold et al., 2011; Shaheen et al., 2011a). Brochard et al. (2011) even described a double calibration technique in which the locator is applied at 0° and 180° of humeral elevation. They modified the 3D positions of anatomical landmarks measured with optical markers between these two postures by linear interpolation in time (Cappello et al., 2005; Zhang, 2002). Recently, also Cereatti et al. (2015) applied a double calibration technique (at start and end of the arm range of motion) in combination with an acromion skin marker cluster in cadaveric specimens. Both studies showed that double calibration improved the estimate of scapular kinematics, whereas single calibration resulted in errors at higher humeral elevation angles (Brochard et al., 2011; Cereatti et al., 2015) since it possibly only corrects for initial scapula angular offsets but not for soft tissue artefacts during the scapular motion.

Therefore, calibration with a scapula locator might correct for offset differences in retraction/protraction and anterior/posterior tilt as observed by van den Noort et al. (van den Noort et al., 2014). Furthermore, double calibration at both low and high humeral elevation angles might be promising to further optimize the measurement of scapular kinematics with IMMS, by correcting for the underestimation of scapular lateral rotation at high elevation angles with a skin-fixed sensor (Meskers et al., 2007). In this way, scapular kinematics measurement with IMMS might be as close as possible to measurement with bone pins, which can be considered to be the ‘golden standard’ (Brochard et al., 2011; McClure et al., 2001). The primary purpose of this study was to evaluate the change in 3D scapular kinematics caused by single and double anatomical calibration with a scapula locator versus standard calibration using sensor alignment with the spina scapulae (ISEO-protocol (Cutti et al., 2008; van den Noort et al., 2014)) in patients with scapular dyskinesis. Single calibration was performed at 0°, while 120° humeral anteflexion (AF) and humeral abduction (AB) was added for double calibration. We hypothesized that double calibration will result in higher scapular lateral rotation angles compared to single calibration for higher humeral elevation angles. The secondary aim of the study was to evaluate the difference in 3D scapular kinematics between static posture (arm elevation of 0°, 30°, 60°, 90° and 120°) and dynamic humeral elevation (movement). This is of interest because Brochard et al. (Brochard et al., 2011) compared their results of double calibration with data obtained in static postures. We hypothesized that scapular lateral rotation is higher during dynamic motion than in static posture (Fayad et al., 2006).

2. Methods 2.1. Patients Ten patients with scapular dyskinesis according to the scapular dyskinesis test (SDT) (McClure et al., 2009; Tate et al., 2009) participated in the study. Underlying shoulder pathologies varied between patients. Patients characteristics are shown in Table 1. For each patient, the most affected shoulder was selected and analyzed. Measurements were performed at the VU University Medical Center (VUmc) in Amsterdam (The Netherlands), department of rehabilitation medicine. The Medical Ethics Committee of the VUmc approved the study protocol. Full written informed consent was obtained from all patients.

Table 1 Patient characteristics. Patient number

Gender

Age

BMI

Side

Diagnosis

SDT

SRQ

CM

1 2 3 4 5 6 7 8 9 10

M M F M M M M M F M

54 55 28 24 43 27 58 47 46 63

23.7 22.8 22.5 23.1 23.9 24.3 24.2 31.6 20.1 24.1

R R L R L R L L R R

Anterior labrum tear Shoulder instability, tendinopathy Posterior impingement Unstable painfull shoulder Glenohumeral luxation SLAP laesion Partial supraspinatus tear Shoulder instability, tendinopathy Tendinitis calcarea Full thickness supraspinatus tear

AF AF AF AF AF AF AF AF AF AF

59 96 94 Unknown 69 82 77 48 71 58

81 95 84 Unknown 86 93 81 45 77 53

normal. AB obvious subtle, AB obvious subtle, AB obvious obvious, AB obvious obvious, AB obvious subtle, AB subtle subtle, AB subtle obvious, AB obvious obvious, AB subtle obvious, AB obvious

M ¼male, F ¼female, BMI¼ Body Mass Index (kg/m2), R¼ right, L ¼left, SDT¼ Scapular Dyskinesis Test, AF ¼ anteflexion, AB¼ abduction, SRQ ¼Shoulder Rating Questionnaire (0–100, a score of 100 implies no complaints), CM ¼ Constant-Murley Shoulder Outcome Score (0–100, a score of 100 implies no complaints).

Please cite this article as: van den Noort, J.C., et al., Measurement of scapular dyskinesis using wireless inertial and magnetic sensors: Importance of scapula calibration. Journal of Biomechanics (2015), http://dx.doi.org/10.1016/j.jbiomech.2015.05.036i

J.C. van den Noort et al. / Journal of Biomechanics ∎ (∎∎∎∎) ∎∎∎–∎∎∎

2.2. Instrumentation Four wireless IMMS sensors (MTw, Xsens Technologies, NL) were placed on the thorax, scapula, upper arm and lower arm of the patients using straps and skin friendly tape, according to the ISEO-protocol described in Cutti et al. (2008); van den Noort et al. (2014). The sensor on the scapula was placed with the x-axis pointing lateral along the cranial edge of the spina scapulae, halfway on the line between acromion and medial border of the scapula, using palpation for correct placement (van den Noort et al., 2014). Another wireless IMMS sensor was fixed to a custom-made scapula locator (transparent polymethylmethacrylate (PMMA) frame with polyoxymethylene (POM) adjustable sliders and pins), see Fig. 1. 2.3. Measurement procedure Assessment of the presence of scapular dyskinesis was done by clinical observation using the Scapular Dyskinesis Test (SDT), in which five repetitions of bilateral, active humeral anteflexion (AF) and humeral abduction (AB) were performed by the patients, as far as possible to a 3-s count using the “thumbs-up” position (thumb pointing lateral or up) (McClure et al., 2009; Tate et al., 2009). Tests were performed while patients grasping dumbbells (1 kg for women and 2 kg for men). An experienced physical therapist rated the dyskinesis as being normal, subtle or obvious (see Table 1) (McClure et al., 2009). Additionally, patients filled in the Shoulder Rating Questionnaire (SRQ) (Vermeulen et al., 2005) to assess global assessment, pain, daily activities, areas for improvement and satisfaction, recreational and athletic activities and work. The Constant-Murley shoulder assessment (Constant and Murley, 1987) was performed by the physical therapist to assess pain, range of motion, strength and function. For both tests, a score of 100 implies no complaints (see Table 1).

3

IMMS were first anatomically calibrated following the ISEO protocol, which includes a static posture (trunk upright, upper arm along the trunk, elbow in 90° flexion), elbow flexion/extension and pro/supination (Cutti et al., 2008). Additionally, the scapula locator was adjusted to and placed on the scapula (the angulus inferior, angulus acromialis and trigonum scapulae, see Fig. 1) and measurements were performed at 0° humeral elevation, and at 30°, 60°, 90° and 120° of active static humeral elevation (elbow fully extended and thumb pointing lateral or up, without additional dumbbells) for both AF and AB. For each angle, the scapula locator was repositioned. The humeral elevation angles were first explored using a goniometer. Elevation higher than 120° was not measured since not all patients were able to elevate their humerus above 120° due to pain or limited range of motion. Subsequently, patients repeatedly performed in total six movement trials with five repetitions each. Three trials consisted of bilateral, active AF in the sagittal plane, whereas the other three trials consisted of bilateral, active AB in the frontal plane (elbow fully extended and thumb pointing up, without additional weight). Maximal humeral elevation was as high as possible without introducing pain or discomfort. The physical therapist determined the rhythm of the elevations by counting, in a same manner to all patients. IMMS signals were recorded with a sample frequency of 75 Hz (IMw_pro_iseos software, 2012, Xsens Technologies, NL). In the analyses, data of the middle three repetitions of all three trials of humeral elevation per subject were analysed and averaged (i.e. nine repetitions for AF and nine repetitions for AB with standard deviations between repetitions on average about 1°). 2.4. Data analysis For the static measurements (0°, 30°, 60°, 90° and 120° of humeral elevation), five time-invariant rotation matrices representing the orientation of the technical frame of the sensor attached to the skin GlRSkinSen and of the sensor on the scapula locator GlRLocator respectively, were obtained. The orientation matrices of the scapula locator were expressed in the thorax segment (for all angular configurations p) using:

⎡ ThR ⎤ ⎡Gl ⎤ T ⎡Gl ⎤ ⎣ Locator ⎦ p = ⎣ R Th⎦ p • ⎣ R Locator ⎦ p

(1)

with T indicating the transpose of the matrix. The orientation matrices of the skin-fixed sensor on the spina scapulae during both static postures and movement trials, calibrated using alignment with the spina scapulae, were obtained from the software using the ISEO-protocol, and also expressed in the thorax segment ([ThRSkin]p and ThRSkin(t), with p being the angular configuration in the static trials and with t being the sample instant of time of the movement trials). For the resting posture (p is 0°) and for the highest static posture (p is 120°), the orientation of the skin-fixed sensor with respect to the locator sensor frames LocatorRSkinSen was calculated, using:

[ Locator R SkinSen]p = [GlR Locator ]Tp • [GlR SkinSen]p

(2)

with p being either p0 or p120, i.e. at 0° and at 120° humeral elevation. For single calibration, [LocatorRSkinSen]p0 at 0° was used to anatomically calibrate the orientation of the skin-fixed sensor during the movement trial, using:

[ ThR Single (t )] = [GlR Th (t )]T •[GlR SkinSen (t )]• [ Locator R SkinSen]p0 Fig. 1. The scapula locator with attached IMMS sensor and IMMS sensor on the spina scapulae.

T

(3)

Gl

with RTh being the orientation of the thorax segment and t being the sample instant of time of the movement trials.

Please cite this article as: van den Noort, J.C., et al., Measurement of scapular dyskinesis using wireless inertial and magnetic sensors: Importance of scapula calibration. Journal of Biomechanics (2015), http://dx.doi.org/10.1016/j.jbiomech.2015.05.036i

J.C. van den Noort et al. / Journal of Biomechanics ∎ (∎∎∎∎) ∎∎∎–∎∎∎

4

For double calibration, interpolation of [LocatorRSkinSen]p between 0° and 120° ([LocatorRSkinSen]p0 and [LocatorRSkinSen]p120) was performed using spherical linear interpolation (based on (Dam et al., 1998; Shoemake, 1985), described in Appendix A) to obtain [LocatorRSkinSen(tp0  p120)]. The orientation of the skin-fixed sensor during the movement trial was then anatomically double calibrated using:

⎡ ThR ⎤ ⎡Gl ⎤T ⎡Gl ⎤ ⎣ Double (t ) ⎦ = ⎣ R Th (t p0 − p120 ) ⎦ • ⎣ R SkinSen t p0 − p120 ⎦

(

T • ⎡⎣ Locator R SkinSen t p0 − p120 ⎤⎦

(

Gl

)

)

during 0°, 30°, 60°, 90°, 120° and 150° (only AB) of humeral elevation of the AF and AB movement trials. Furthemore, pairwise comparisons were performed between (4) Skin and Locator (scapula locator) at 0°, 30°, 60°, 90° and 120° of static humeral elevation for both AF and AB to estimate offsets; and (5) Locator of the static trials and Double of the movement trials at 0°, 30°, 60°, 90° and 120° of humeral elevation of AF and AB to estimate the effect of movement versus static posture. A P-value of o0.05 was considered to be significant.

(4)

with RTh being the orientation of the thorax segment and tp0  p120 being the sample instant of time of the movement trials of humeral elevation between 0° and 120°. For elevations higher than 120° [LocatorRSkinSen]p120 was used. Single and double calibrations were performed for AF and AB separately. [ThRLocator]p, [ThRSkin]p, [ThRSkin(t)], [ThRSingle(t)] and [ThRDouble(t)] were decomposed into scapulo-thoracic angles using the YZX Euler sequence (Y: retraction/protraction, Z: medio/lateral rotation, X: anterior/posterior tilt). Decomposition of the humerothoracic orientation matrix to 3D angles was depending on movement in the sagittal (AF) or frontal (AB) plane (sagittal XZY, frontal ZXY) (Cutti et al., 2008; van den Noort et al., 2014). Linear Mixed Models (SPSS software), including pairwise comparisons with Bonferroni correction were used to determine the mean difference, significance and 95% confidence intervals in 3D angles between (1) Skin (calibrated skin sensor from software using alignment with spina scapulae (ISEO)) and Single (single calibration with locator at 0°), (2) Skin and Double (double calibration with locator at 0° and 120°), and (3) Single and Double

3. Results Both single and double calibration resulted in a significant increase of scapular anterior tilt for all elevation angles during dynamic anteflexion and abduction (AF 17–21°; AB 14–30°) compared to the skin-fixed sensor calibrated using alignment (ISEO) (see Fig. 2 and Table 2). Both single and double calibration showed 47° less protraction during abduction with respect to the skinfixed sensor, however these differences were not significant (P 40.05). For abduction at higher elevation angles, double calibration resulted in increased lateral rotation angles compared to the skin-fixed sensor (8–15° at Z 60° abduction), although these differences were not significant (P4 0.09). Compared to single calibration, double calibration did not show a significant higher lateral rotation or anterior tilt for anteflexion and abduction. For elevation angles Z90° humeral abduction differences of 10–14° between single and double calibration were observed, however these differences were not significant (P 40.06). For 150° abduction, 8° increased anterior tilt was measured with double

Fig. 2. Scapular kinematics (mean) during humeral anteflexion and abduction movements of 10 patients with scapular dyskinesis. Shown are the mean values from the sensor on the skin aligned with the spina scapulae (ISEO; red solid line), after single calibration at 0° with the scapula locator (blue dashed line) and after double calibration at 0° and 120° with the scapula locator (black dotted line). Standard deviations (SD) are not shown to enhance visibility. For re/protraction SDs vary between 15–27°, for medio/ lateral rotation between 5–13° and for ant/posterior tilt between 2–12° (see also Figs. 3 and 4). Table 2 includes mean differences, standard errors and confidence intervals. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

Please cite this article as: van den Noort, J.C., et al., Measurement of scapular dyskinesis using wireless inertial and magnetic sensors: Importance of scapula calibration. Journal of Biomechanics (2015), http://dx.doi.org/10.1016/j.jbiomech.2015.05.036i

J.C. van den Noort et al. / Journal of Biomechanics ∎ (∎∎∎∎) ∎∎∎–∎∎∎

5

Table 2 Single and double calibration compared to the aligned sensor on the skin (spina scapulae) during the movement trials (N ¼ 10). Humeral anteflexion [deg]

Mean difference [deg]

0

Ant/posterior tilt

30

Ant/posterior tilt

60

Ant/posterior tilt

90

Ant/posterior tilt

120

Ant/posterior tilt

Humeral abduction 60 90

Medio/lateral rotation Medio/lateral rotation

120

Medio/lateral rotation

150

Medio/lateral rotation

0

Re/protraction

30

Re/protraction

60

Re/protraction

90

Re/protraction

120

Re/protraction

150

Re/protraction

0

Ant/posterior tilt

30

Ant/posterior tilt

60

Ant/posterior tilt

90

Ant/posterior tilt

120

Ant/posterior tilt

150

Ant/posterior tilt

n

Std. error [deg]

Sig.

95% Confidence interval for difference Lower bound [deg] Upper bound [deg]

Single vs. Skin Double vs. Skin Single vs. Skin Double vs. Skin Single vs. Skin Double vs. Skin Single vs. Skin Double vs. Skin Single vs. Skin Double vs. Skin

 18.1  17.5  17.9  18.2  17.4  18.8  17.1  19.2  19.3  20.5

1.7 1.8 2.0 2.0 2.5 2.5 3.5 3.5 5.3 5.3

0.00n 0.00n 0.00n 0.00n 0.00n 0.00n 0.00n 0.00n 0.01n 0.01n

 23.1  22.8  24.0  24.2  24.8  26.2  27.5  29.6  35.4  36.6

 13.0  12.3  11.9  12.2  10.0  11.4  6.8  8.8  3.2  4.4

Double vs. Skin Double vs. Skin Double vs. Single Double vs. Skin Double vs. Single Double vs. Skin Double vs. Single Single vs. Skin Double vs. Skin Single vs. Skin Double vs. Skin Single vs. Skin Double vs. Skin Single vs. Skin Double vs. Skin Single vs. Skin Double vs. Skin Single vs. Skin Double vs. Skin Single vs. Skin Double vs. Skin Single vs. Skin Double vs. Skin Single vs. Skin Double vs. Skin Single vs. Skin Double vs. Skin Single vs. Skin Double vs. Skin Single vs. Skin Double vs. Skin Double vs. Single

7.8 9.5 9.8 11.7 11.8 15.2 14.4  7.1  7.1  7.8  7.6  7.7  7.9  7.6  9.0  7.3  11.1  9.5  12.1  16.4  16.6  16.9  19.5  14.8  19.7  13.7  20.5  17.1  23.3  23.7  32.1  8.4

3.4 3.4 3.4 4.1 4.1 10.0 10.1 11.4 11.4 9.7 9.7 9.7 9.7 9.9 10.1 10.1 10.1 22.9 24.8 2.5 2.5 2.8 2.9 3.6 3.6 4.0 4.0 4.2 4.2 8.3 8.8 8.8

0.29 0.08 0.06 0.07 0.06 0.46 0.53 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 0.00n 0.00n 0.00n 0.00n 0.00n 0.00n 0.01n 0.00n 0.00n 0.00n 0.05n 0.01n 1.00

 2.4  0.6  0.3  0.5  0.4  12.9  13.7  41.3  41.4  36.8  36.6  36.8  37.0  37.2  38.6  37.4  41.2  78.7  86.8  23.9  24.1  25.3  28.3  25.7  30.6  25.7  32.5  29.7  36.0  47.2  57.0  33.3

17.9 19.6 19.9 23.9 23.9 43.4 42.5 27.2 27.1 21.2 21.4 21.4 21.7 22.0 20.5 22.8 18.9 59.7 62.6  8.9  9.1  8.4  10.8  4.1  8.9  1.8  8.6  4.5  10.7  0.2  7.2 16.5

Po 0.05.

calibration compared to single calibration, but also this difference was significant. In static posture, the scapula locator showed a significant increase of scapular anterior tilt for all static humeral elevation angles in both anteflexion and abduction postures compared to the aligned skin-fixed sensor (AF 14–21°; AB 16–22°) (see Fig. 3 and Table 3). Increased lateral rotation ( 47°) was measured with the locator during 120° anteflexion and during 460° abduction, which was only significant for a humeral abduction of 90° (10°, P ¼0.03). Protraction was not significantly different (P 40.05). When comparing the dynamic movement trials (from double calibration) with the static postures (from the locator), higher lateral rotation angles (4 8°) were observed during Z60° anteflexion and Z30° abduction (P 40.05). The differences were only significant for 90° anteflexion (14°, P¼ 0.03) (see Fig. 4 and Table 4). 4. Discussion Calibration of IMMS with a scapula locator resulted in measurement of increased anterior tilt during both humeral

anteflexion (AF) and abduction (AB) compared to the skin-fixed sensor calibrated using alignment (ISEO-protocol). The measurement of re/protraction was not significantly influenced by use of a scapula locator. In contrast to our hypothesis, double calibration did not result in higher scapular lateral rotation angles compared to single calibration during AF movement. Only in static AF posture, increased lateral rotation of 8° was measured at 120° of elevation compared to the skin-fixed sensor, but this was not significant. However, during static AB posture, significant increased lateral rotation of 10° was measured at 90° of elevation compared to the skin-fixed sensor. Furthermore, during AB movement, double calibration showed increased lateral rotation angles of 10–14° at 90–150° humeral elevation compared to single calibration, being close to significance. Confirming our hypothesis, increased lateral rotation of 8–14° was measured during dynamic motion with respect to static elevation of the humerus, although only significantly different for 90° AF. In a previous study (van den Noort et al., 2014) we observed that using alignment of the IMMS sensor on the skin with the spina scapulae in combination with a static upright posture for anatomical calibration (ISEO), the measure of anterior/posterior

Please cite this article as: van den Noort, J.C., et al., Measurement of scapular dyskinesis using wireless inertial and magnetic sensors: Importance of scapula calibration. Journal of Biomechanics (2015), http://dx.doi.org/10.1016/j.jbiomech.2015.05.036i

J.C. van den Noort et al. / Journal of Biomechanics ∎ (∎∎∎∎) ∎∎∎–∎∎∎

6

tilt may be very subjective to offset. The results of both single and double calibration clearly show that the scapula locator results in offset differences in anterior/posterior tilt when using IMMS (14– 32°). Offsets that were observed in re/protraction angles when using ISEO seem not to be corrected by use of a scapula locator, since single or double calibration did not significantly change the re/protraction angles. For AB, a decrease in protraction angles of 7–

14° was measured, which is in between the SEM and the smallest detectable difference (SDD) of inter-observer reliability as observed in van den Noort et al. (2014) (SEM 6–8°, SDD 13–21°). However, due to high standard deviations (SD) within the patient population and high standard errors (SE) these differences are not significant (SD 15–27°, SE of the difference 9–21°). This is in line with a review paper by De Baets et al. that shows high

Fig. 3. Scapular angles (mean and standard deviations) during static humeral anteflexion and abduction postures of 10 patients with scapular dyskinesis, measured with from the sensor on the skin aligned with the spina scapulae (ISEO; red dots) and measured with the scapula locator (blue stars). (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

Table 3 The scapula locator compared to the aligned sensor on the skin (spina scapulae) in static posture (N ¼ 10). Humeral anteflexion [deg]

Mean difference [deg]

Std. error [deg]

Sig.

95% Confidence interval for difference Lower bound [deg] Upper bound [deg]

120 0 30 60 90 120

Medio/lateral rotation Ant/posterior tilt Ant/posterior tilt Ant/posterior tilt Ant/posterior tilt Ant/posterior tilt

7.8  17.4  17.2  14.2  15.6  20.7

3.4 1.7 2.0 2.6 3.6 4.5

0.27 0.00n 0.00n 0.00n 0.00n 0.00n

 2.3  22.5  23.3  21.8  26.2  34.4

17.9  12.3  11.2  6.7  4.9  7.1

Humeral abduction 90 120 60 90 120 0 30 60 90 120

Re/protraction Re/protraction Medio/lateral rotation Medio/lateral rotation Medio/lateral rotation Ant/posterior tilt Ant/posterior tilt Ant/posterior tilt Ant/posterior tilt Ant/posterior tilt

 10.0  8.4 7.9 10.3 12.0  16.3  18.4  17.7  18.7  22.1

9.4 9.5 3.2 3.2 4.1 2.3 2.8 3.4 3.7 4.2

1.00 1.00 0.20 0.03n 0.06 0.00n 0.00n 0.00n 0.00n 0.00n

 38.0  36.7  1.8 0.8  0.2  23.2  26.7  27.9  30.0  34.8

18.1 19.9 17.5 19.9 24.2  9.5  10.2  7.6  7.5  9.5

n

Po 0.05.

Please cite this article as: van den Noort, J.C., et al., Measurement of scapular dyskinesis using wireless inertial and magnetic sensors: Importance of scapula calibration. Journal of Biomechanics (2015), http://dx.doi.org/10.1016/j.jbiomech.2015.05.036i

J.C. van den Noort et al. / Journal of Biomechanics ∎ (∎∎∎∎) ∎∎∎–∎∎∎

inconsistency in scapular transversal plane movement in healthy subjects (De Baets et al., 2013), whereas this might be even more exaggerated in our patient population due to the different shoulder pathologies. Several studies showed that the pattern of re/protraction is very inconsistent in persons with primary shoulder disorders (Borstad and Ludewig, 2002; De Baets et al., 2013; Ludewig and Cook, 2000; Matias and Pascoal, 2006). When aiming to treat scapular dyskinesis within the physical therapy setting, enhancement of the IMMS measurement of lateral rotation and anterior tilt is crucial, since variable changes in these angles are often seen in patient populations (Borstad and Ludewig, 2002; De Baets et al., 2013; Ludewig and Cook, 2000; Ludewig and Reynolds, 2009; Lukasiewicz et al., 1999; McClure et al., 2004, 2006; Ogston and Ludewig, 2007; Struyf et al., 2011). For example, it has been reported that patients with instability show decreased lateral rotation and decreased anterior tilt (De Baets et al., 2013; Lukasiewicz et al., 1999; Ogston and Ludewig, 2007; Struyf et al., 2011), whereas in patients with impingement or rotator cuff tears both decreased and increased lateral rotation angles and posterior tilt angles have been reported (Balke et al., 2013; De Baets et al.,

7

2013; Lukasiewicz et al., 1999; McClure et al., 2006; Struyf et al., 2011). The use of a single calibration clearly changes the anterior tilt with an initial offset of more than 14° in both static and dynamic conditions (Tables 2 and 3), which is beyond the SDD as measured in van den Noort et al. (2014). The additional value of double calibration to e.g. overcome the effect of soft tissue artefacts at higher humeral elevation angles is less obvious. An increase of 8° anterior tilt of double versus single calibration is seen for AB at 150°, however this is not significant. Furthermore, our study population shows high variability of lateral rotation angles of the scapula at humeral elevation above 60° (SD 6–13°, SE of difference 3–9°). Similar to the re/protraction angles, this may be due to the heterogeneous pattern of movement in a patient population (De Baets et al., 2013). However, in contrast to re/ protraction, the differences in lateral rotation between the skinfixed sensor, single and double calibration (10–14°) for AB Z 90° are most of the time above the SDD as reported in (van den Noort et al., 2014) (i.e. 7–9°). Therefore, differences might be relevant, indicating that double calibration causes measurement of higher lateral rotation angles at high elevation, therewith correcting the

Fig. 4. Scapular angles (mean and standard deviations) during static (scapula locator, blue dots) and dynamic (using double calibration, green plusses) humeral anteflexion and abduction postures of 10 patients with scapular dyskinesis. (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

Table 4 The scapula locator in static posture compared to the double calibration during the movement trials (N ¼ 10). Humeral anteflexion [deg] 60 90 120 Humeral abduction 30 60 90 120 n

Medio/lateral rotation Medio/lateral rotation Medio/lateral rotation Medio/lateral Medio/lateral Medio/lateral Medio/lateral

rotation rotation rotation rotation

Mean difference [deg]

Std. error [deg]

Sig.

95% Confidence interval for difference Lower bound [deg] Upper bound [deg]

 10.1  13.6  9.2

4.2 4.4 3.7

0.20 0.03n 0.18

 22.5  26.5  20.3

2.2  0.7 1.9

 8.6  9.1  8.7  8.9

3.6 3.4 3.4 4.1

0.21 0.11 0.15 0.35

 19.3  19.2  18.8  21.1

2.0 1.1 1.4 3.2

Po 0.05.

Please cite this article as: van den Noort, J.C., et al., Measurement of scapular dyskinesis using wireless inertial and magnetic sensors: Importance of scapula calibration. Journal of Biomechanics (2015), http://dx.doi.org/10.1016/j.jbiomech.2015.05.036i

J.C. van den Noort et al. / Journal of Biomechanics ∎ (∎∎∎∎) ∎∎∎–∎∎∎

8

underestimation with a skin-fixed sensor and with single calibration (Brochard et al., 2011; Meskers et al., 2007). In particular for patients with high body mass index (BMI), soft tissue artefacts due to fat tissue or muscle bulging may effect the movement of the skin with respect to the underlying scapular bone. For these patients, the use of double calibration might improve the measurement of scapular lateral rotation above 90° of humeral elevation. In the current study, all patients except one had a BMI below 25 (Table 1), therefore this assumption could not further be evaluated. Brochard et al. (2011) showed that when using optical markers for scapular kinematic measurement, double calibration with a scapula locator mainly affected the measure of anterior tilt and lateral rotation at high elevation angles. Also Cereatti et al. (2015) showed that double calibration improved scapular kinematics (up to 14°). We did not find a significant effect of double calibration for anterior tilt and lateral rotation, although increased lateral rotation was measured above 90° AB (10–14°, P 40.06). The different findings might be due to the study population (young healthy subjects with low variability (Brochard et al., 2011) and cadaveric specimen without shoulder damage (Cereatti et al., 2015) versus patients with shoulder pathology with high variability). Furthermore, in the set-up of (Brochard et al., 2011) and (Cereatti et al., 2015) the skin-fixed markers were placed on a different position on the scapula, i.e. not aligned with the spina scapulae but on the acromion (as proposed by van Andel et al. (2009) for optical markers). Brochard et al. compared their double calibration data with static palpation data. This approach may be similar to our comparison between scapula locator in static posture and double calibration during dynamic movement (Fig. 4 and Table 4). Scapular lateral rotation angles are different during motion compared to static postures (Fayad et al., 2006). Increased lateral rotation at 90° of AF movement was observed. This might be mainly due to muscle activity (e.g. from the deltoid Veeger and van der Helm (2007)) causing movement artefacts. Cereatti et al. used cadaveric specimen to evaluate double calibration during movement, however soft tissue artefacts during motion might have been less in comparison to living humans due to the lack of muscular contraction (Cereatti et al., 2015). A limitation of the current study is the rather small heterogeneous patient group. Follow-up studies are necessary to evaluate scapular dyskinesis measured with IMMS and scapula locator in a more homogeneous patient group, e.g. limited to patients with shoulder instability. This should also include the evaluation of interventions, such as a physical therapy exercise program (shoulder muscle coordination or strength training) or even surgical intervention. Finally double calibration was performed at an elevation angle of 120°, whereas Brochard et al. (2011) used an elevation angle of 180°. However, high elevation angles are often not feasible, since, at least in our group, not all patients were able to elevate above 120° due to pain and limitations in range of motion. Also most functional movements will not exceed an elevation angle of 120°. Double calibration at 120° does therefore not limit a meaningful assessment.

application of double calibration is required, e.g. in patients with high body mass index.

Conflict of interest statement The authors state that there are no financial and personal relationships with other people or organizations that could inappropriately influence (bias) their work.

Acknowledgment The authors would like to thank Julien Pronk for design and manufacture of the scapula locator.

Appendix A For double calibration, interpolation of [LocatorRSkinSen]p between 0° and 120° ([LocatorRSkinSen]p0 and [LocatorRSkinSen]p120) was performed using spherical linear interpolation (slerp (Shoemake, 1985; Dam et al., 1998)) to obtain [LocatorRSkinSen(tp0-p120)]. Therefore, the rotation matrices have to be transformed to quaternions, using (Kuipers, 2002; Shoemake, 1985):

⎫ 1 ⋅ 1 + R1,1 + R2,2 + R 3,3 ⎬ scalar part ∈ R ⎭ 2 ⎫ 1 R2,3 − R 3,2 ⎪ q1 = 4 ⎪ q0 ⎪ 1 ⎪ ⎪ − R R 3,1 1,3 ⎬ vector part ∈ R 3 q2 = 4 ⎪ q0 ⎪ 1 R1,2 − R2,1 ⎪ 4 ⎪ q3 = ⎪ q0 ⎭

q0 =

(

)

(

)

(

)

(5)

The following equation was used to perform the spherical linear interpolation (Shoemake, 1985; Dam et al., 1998):

qm (t ) =

qa sin ((1 − ts (t ) ) θ ) + qb sin (ts (t ) θ ) sin (θ )

qa is quaternion at p0 , qb is quaternion at p120 , qm is interpolated quaternion p0 −p120 ts (t ) is a scalar between 0 and 1 over time t , related to the number of samples that have to be interpolated between 0° and 120° humeral elevation θ is half the rotation angle between qa and qb with: cos (θ ) = qa 0⋅qb0 + qa1⋅qb1 + qa2⋅qb2 + qa 3⋅qb3

5. Conclusion

ifcos (θ ) < 0, qa = − qa, qb

Calibration with a scapula locator when applying IMMS is necessary, because measures of scapular anterior tilt are grossly underestimated with the ISEO-protocol. For shoulder movements that exceed 90° elevation, a double calibration prevents small but relevant underestimation of lateral rotation angles of the scapula on top of correction of initial offsets by single calibration. Enhanced evaluation is needed to evaluate whether broader

= − qb, cos (θ ) = − cos (θ ) since θ should be between 0° and 90° sin (θ ) =

1 − cos2 (θ )

(6)

Please cite this article as: van den Noort, J.C., et al., Measurement of scapular dyskinesis using wireless inertial and magnetic sensors: Importance of scapula calibration. Journal of Biomechanics (2015), http://dx.doi.org/10.1016/j.jbiomech.2015.05.036i

J.C. van den Noort et al. / Journal of Biomechanics ∎ (∎∎∎∎) ∎∎∎–∎∎∎

References Balke, M., Schmidt, C., Dedy, N., Banerjee, M., Bouillon, B., Liem, D., 2013. Correlation of acromial morphology with impingement syndrome and rotator cuff tears. Acta Orthop. 84, 178–183. Barnett, N.D., Duncan, R.D., Johnson, G.R., 1999. The measurement of three dimensional scapulohumeral kinematics–a study of reliability. Clin. Biomech. 14, 287–290. Borstad, J.D., Ludewig, P.M., 2002. Comparison of scapular kinematics between elevation and lowering of the arm in the scapular plane. Clin. Biomech. 17, 650–659. Brochard, S., Lempereur, M., Remy-Neris, O., 2011. Double calibration: an accurate, reliable and easy-to-use method for 3D scapular motion analysis. J. Biomech. 44, 751–754. Cappello, A., Stagni, R., Fantozzi, S., Leardini, A., 2005. Soft tissue artifact compensation in knee kinematics by double anatomical landmark calibration: performance of a novel method during selected motor tasks. IEEE Trans. Biomed. Eng. 52, 992–998. Cereatti, A., Rosso, C., Nazarian, A., DeAngelis, J.P., Ramappa, A.J., Della Croce, U., 2015. Scapular motion tracking using acromion skin marker cluster: in vitro accuracy assessment. J. Med. Biol. Eng. 35, 94–103. Constant, C.R., Murley, A.H., 1987. A clinical method of functional assessment of the shoulder. Clin. Orthop. Relat. Res., 160–164. Cutti, A.G., Giovanardi, A., Rocchi, L., Davalli, A., Sacchetti, R., 2008. Ambulatory measurement of shoulder and elbow kinematics through inertial and magnetic sensors. Med. Biol. Eng. Comput. 46, 169–178. Dam, E.B., Koch, M., Lillholm, M., 1998. Quaternions, Interpolation and Animation. Interpolation of rotation. University of Copenhagen, Copenhagen, pp. 42–48. De Baets, L., Jaspers, E., Desloovere, K., Van Deun, S., 2013. A systematic review of 3D scapular kinematics and muscle activity during elevation in stroke subjects and controls. J. Electromyogr. Kinesiol. 23, 3–13. de Groot, J.H., 1997. The variability of shoulder motions recorded by means of palpation. Clin. Biomech. 12, 461–472. Fayad, F., Hoffmann, G., Hanneton, S., Yazbeck, C., Lefevre-Colau, M.M., Poiraudeau, S., Revel, M., Roby-Brami, A., 2006. 3-D scapular kinematics during arm elevation: effect of motion velocity. Clin. Biomech. 21, 932–941. Johnson, G.R., Stuart, P.R., Mitchell, S., 1993. A method for the measurement of 3-dimensional scapular movement. Clin. Biomech. 8, 269–273. Kibler, W.B., Ludewig, P.M., McClure, P., Uhl, T.L., Sciascia, A., 2009. Scapular summit 2009: introduction. July 16, 2009, Lexington, Kentucky. J. Orthop. Sports Phys. Ther. 39, A1–A13. Kibler, W.B., Sciascia, A., 2010. Current concepts: scapular dyskinesis. Br. J. Sports Med. 44, 300–305. Kuipers, J.B., 2002. Quaternions and Rotation Sequences. Princeton University Press, Princeton, New Jersey. Ludewig, P.M., Cook, T.M., 2000. Alterations in shoulder kinematics and associated muscle activity in people with symptoms of shoulder impingement. Phys. Ther. 80, 276–291. Ludewig, P.M., Reynolds, J.F., 2009. The association of scapular kinematics and glenohumeral joint pathologies. J. Orthop. Sports Phys. Ther. 39, 90–104. Lukasiewicz, A.C., McClure, P., Michener, L., Pratt, N., Sennett, B., 1999. Comparison of 3-dimensional scapular position and orientation between subjects with and without shoulder impingement. J. Orthop. Sports Phys. Ther. 29, 574–583. Matias, R., Pascoal, A.G., 2006. The unstable shoulder in arm elevation: a threedimensional and electromyographic study in subjects with glenohumeral instability. Clin. Biomech. 21 (Suppl. 1), S52–S58. McClure, P., Tate, A.R., Kareha, S., Irwin, D., Zlupko, E., 2009. A clinical method for identifying scapular dyskinesis, part 1: reliability. J. Athl. Train. 44, 160–164. McClure, P.W., Bialker, J., Neff, N., Williams, G., Karduna, A., 2004. Shoulder function and 3-dimensional kinematics in people with shoulder impingement syndrome before and after a 6-week exercise program. Phys. Ther. 84, 832–848.

9

McClure, P.W., Michener, L.A., Karduna, A.R., 2006. Shoulder function and 3-dimensional scapular kinematics in people with and without shoulder impingement syndrome. Phys. Ther. 86, 1075–1090. McClure, P.W., Michener, L.A., Sennett, B.J., Karduna, A.R., 2001. Direct 3-dimensional measurement of scapular kinematics during dynamic movements in vivo. J. Shoulder Elb. Surg. 10, 269–277. Meskers, C.G., van de Sande, M.A., de Groot, J.H., 2007. Comparison between tripod and skin-fixed recording of scapular motion. J. Biomech. 40, 941–946. Ogston, J.B., Ludewig, P.M., 2007. Differences in 3-dimensional shoulder kinematics between persons with multidirectional instability and asymptomatic controls. Am. J. Sports Med. 35, 1361–1370. Parel, I., Cutti, A.G., Fiumana, G., Porcellini, G., Verni, G., Accardo, A.P., 2012. Ambulatory measurement of the scapulohumeral rhythm: intra- and interoperator agreement of a protocol based on inertial and magnetic sensors. Gait Posture 35, 636–640. Parel, I., Cutti, A.G., Kraszewski, A., Verni, G., Hillstrom, H., Kontaxis, A., 2014. Intraprotocol repeatability and inter-protocol agreement for the analysis of scapulohumeral coordination. Med. Biol. Eng. Comput. 52, 271–282. Picerno, P., Cereatti, A., Cappozzo, A., 2008. Joint kinematics estimate using wearable inertial and magnetic sensing modules. Gait Posture 28, 588–595. Prinold, J.A., Shaheen, A.F., Bull, A.M., 2011. Skin-fixed scapula trackers: a comparison of two dynamic methods across a range of calibration positions. J. Biomech. 44, 2004–2007. Roren, A., Fayad, F., Poiraudeau, S., Fermanian, J., Revel, M., Dumitrache, A., Gautheron, V., Roby-Brami, A., Lefevre-Colau, M.M., 2013. Specific scapular kinematic patterns to differentiate two forms of dynamic scapular winging. Clin. Biomech. 28, 941–947. Shaheen, A.F., Alexander, C.M., Bull, A.M., 2011a. Effects of attachment position and shoulder orientation during calibration on the accuracy of the acromial tracker. J. Biomech. 44, 1410–1413. Shaheen, A.F., Alexander, C.M., Bull, A.M., 2011b. Tracking the scapula using the scapula locator with and without feedback from pressure-sensors: a comparative study. J. Biomech. 44, 1633–1636. Shoemake, K., 1985. Animating rotation with quaternion curves. In: Proceedings of Association for Computing Machinery 19(3), San Francisco, USA. Struyf, F., Nijs, J., Baeyens, J.P., Mottram, S., Meeusen, R., 2011. Scapular positioning and movement in unimpaired shoulders, shoulder impingement syndrome, and glenohumeral instability. Scand. J. Med. Sci. Sports 21, 352–358. Tate, A.R., McClure, P., Kareha, S., Irwin, D., Barbe, M.F., 2009. A clinical method for identifying scapular dyskinesis, part 2: validity. J. Athl. Train. 44, 165–173. van Andel, C.J., van, H.K., Eversdijk, M., Veeger, D., Harlaar, J., 2009. Recording scapular motion using an acromion marker cluster. Gait Posture 29, 123–128. van den Noort, J.C., Wiertsema, S.H., Hekman, K.M., Schonhuth, C.P., Dekker, J., Harlaar, J., 2014. Reliability and precision of 3D wireless measurement of scapular kinematics. Med. Biol. Eng. Comput. 52, 921–931. Veeger, H.E., van der Helm, F.C., 2007. Shoulder function: the perfect compromise between mobility and stability. J. Biomech. 40, 2119–2129. Vermeulen, H.M., Boonman, D.C., Schuller, H.M., Obermann, W.R., van Houwelingen, H.C., Rozing, P.M., Vliet Vlieland, T.P., 2005. Translation, adaptation and validation of the Shoulder Rating Questionnaire (SRQ) into the Dutch language. Clin. Rehabil. 19, 300–311. Warner, J.J., Micheli, L.J., Arslanian, L.E., Kennedy, J., Kennedy, R., 1992. Scapulothoracic motion in normal shoulders and shoulders with glenohumeral instability and impingement syndrome. a study using Moire topographic analysis. Clin. Orthop. Relat. Res., 191–199. Xsens Technologies B.V., (2011). Motion Tracker Test & Calibration Certificate. The Netherlands. Zhang, X., 2002. Deformation of angle profiles in forward kinematics for nullifying end-point offset while preserving movement properties. J. Biomech. Eng. 124, 490–495.

Please cite this article as: van den Noort, J.C., et al., Measurement of scapular dyskinesis using wireless inertial and magnetic sensors: Importance of scapula calibration. Journal of Biomechanics (2015), http://dx.doi.org/10.1016/j.jbiomech.2015.05.036i