Test-retest reliability and validity of a motion capture (MOCAP) system for measuring thoracic and lumbar spinal curvatures and sacral inclination in the sagittal plane

Abstract

OBJECTIVE:

To evaluate the test-retest reliability and validity of the MOCAP system for measuring spinal sagittal thoracic and lumbar curvatures and sacral inclination in a standing posture.

METHODS:

Twenty-five male adults were evaluated on lateral standing radiographs. The thoracic and lumbar curvatures were calculated by Harrison’s posterior tangent method. The sacral inclination was defined as the angle between the tangent line of the sacral and vertical plane. In addition, MOCAP was used to calculate the spinal curvatures and sacral inclination.

RESULTS:

The thoracic and lumbar spine and sacral inclination demonstrated excellent reliability, with mean ICCs levels greater than 0.980 and low CVs (mean: 2.15%). Systematic biases were not significant and were very near 0, and the mean standard errors were 0.257 ${}^{\circ}$ . ANOVA of the radiographic and MOCAP measures did not report any statistically significant differences in the comparisons. The systematic biases and mean random errors were lower than 1 ${}^{\circ}$ , with CVs lower than 5% and mean ICCs higher than 0.90 between sessions.

CONCLUSION:

The MOCAP system delivered consistently reliable and valid results for standing curvatures compared with a radiographic technique. This system could be used with confidence in research and clinical environments for sagittal spinal curvature measurements.

Keywords

Optitrack kyphosis lordosis spine posture

1. Introduction

The spine is the main axis of the body, and it has several physiological sagittal curvatures. Thoracic kyphosis is the sagittal plane curvature between the T1 and T12 vertebral bodies, and lumbar lordosis is the sagittal plane curvature between L1 and L5. In a standing posture, thoracic curvatures have been classified by Mejia et al. [1] as thoracic hypokyphosis with values below 20 ${}^{\circ}$ , as hyperkyphosis with values above 45 ${}^{\circ}$ and as neutral kyphosis with values between 20 ${}^{\circ}$ and 40 ${}^{\circ}$ . For lumbar curvature, values below 20 ${}^{\circ}$ have been considered hypolordotic, values above 40 ${}^{\circ}$ have been considered hyperlordotic and values between 20 ${}^{\circ}$ and 40 ${}^{\circ}$ are considered neutral [2].

Alterations of spinal physiological curvatures in the sagittal plane have been associated with a predisposition to spinal disorders due to an increase of intradiscal pressure [3], viscoelastic deformation of lumbar tissues [4] and lower back pain events [5]. Evaluating spinal curvatures is required in clinical settings and research studies.

Radiographic methods are the “gold standard” for measuring spinal posture and skeletal alignment [6, 7]. An important limitation of this method is the relative dose of radiation required, which limits its use for postural assessment. Several studies have shown the usability of other techniques or instruments without radiation to evaluate spinal posture or alignment and sacral inclination. The most used techniques and instruments include Debrunner’s kyphometer [8], an arcometer [9], an inclinometer [10], a flexi curve ruler [11], computerized tomography scanning [12], the Moire projection technique [13], the 3-space Fastrack [14], the SpinalMouse [15], the Spinal wheel [16], and photogrammetry [17]. In their systematic review of the reliability and validity of non-radiographic methods of thoracic kyphosis measurement, Barrett et al. [18] reported that there is much evidence of the reliability and validity of the majority of these techniques or instruments. However, further reliability and validity studies are required to strengthen the evidence for other measurement methods. The optoelectronic motion capture (MOCAP) system is the gold standard for measuring human kinematics in clinical studies [19]. To the best of our knowledge, no previous studies have evaluated the reliability and validity of the MOCAP system for measuring spinal sagittal curvatures and sacral inclination. The aim of this study was to evaluate the test-retest reliability and validity of the MOCAP system for measuring spinal sagittal thoracic and lumbar curvatures and sacral inclination in a standing posture.

2. Methods

A total of twenty-five asymptomatic male adults without spinal pathology (the mean values were the following: age, 21.32 $\pm$ 3.12 years; height, 1.76 $\pm$ 0.05 m; body mass, 69.76 $\pm$ 7.22 kg) voluntarily participated in the study. The participants were excluded if they suffered from back pain; had incurred any spinal injury or had any known structural spinal, hip or pelvic disorder; or had any contraindication for radiographic exposure. The participants were instructed to avoid strenuous physical activity 72 h before the study. An institutional ethics committee approved the study, which was performed in accordance with the Declaration of Helsinki. The participants signed written consent forms before entering the study.

Initially, to evaluate the validity of the MOCAP system, a radiographic protocol was standardized for all the study participants. These data were used as the gold standard. For each participant, a standing right lateral radiograph involving the spine and pelvis was obtained with a Carestream DRX-Evolution System (Carestream Health, Inc., Rochester, NY) situated 1.20 m from the participant. The participants were instructed to stand in a comfortable position with their hips and knees fully extended and with their arms raised horizontally forward with their forearms resting on a support (Fig. 1).

Figure 1.

Lateral radiographs of the spine and pelvis were made with the subject in a controlled standing position. The forearms were placed on a support and the patient was asked to stand in a comfortable but erect posture.

The Carestream Image Suite V.4 (Carestream Health,

Inc., Rochester, NY) was used to assess the spinal parameters of the participants. All the computer measurements were performed by the same experienced radiologist. The Harrison’s posterior tangent method was used to compute the spinal parameters, which has been previously described [6, 7, 20]. Harrison’s posterior tangent method uses the posterior-superior and posterior-inferior corners of each vertebral body to create tangent lines. In this study, the absolute angles between T1–T12 (thoracic kyphosis angle – alpha1) and L5–S1 (lumbar lordosis angle – alpha2) were calculated. The sacral inclination was defined as the angle between the tangent line of the sacral and the vertical plane (sacral inclination – alpha3) (Fig. 2).

Figure 2.

Spherical reflective markers placement and angle definition.

Six spherical reflective markers (B&L Engineering, Tustin, CA, USA) were attached to specific anatomic landmarks of the subjects in a standing posture. One marker was placed at the spinous process of both the first and third thoracic vertebrae (V1 and V2), one marker was placed at both the spinous process of the eleventh thoracic and the first lumbar vertebrae (V3 and V4) and one marker was placed at both the spinous process of the fourth lumbar vertebrae and the second sacral vertebrae (V5 and V6). In the sagittal plane, the angle between the segments of the spine defined by the markers V1–V2 and V3–V4 was used to determine the thoracic kyphosis angle (alpha1), the angle between the segments defined by the markers V3–V4 and V5–V6 was used to determine the lumbar lordosis angle (alpha2), whereas the angle between the segments defined by the markers V5–V6 and vertical plane was used to determine the sacral inclination (alpha3).

A schematic representation of the anatomical attached landmark, as well as their respective angles calculated, are shown in Fig. 2.

Prior to videotaping, each participant was encouraged to adopt the same posture as that used during the radiographic evaluation. A system consisting of 10-camera motion capture (Flex 3, Optritrack, Natural Point, OR, USA) calibrated according to the manufacturer’s protocol recorded each participant for 10 seconds at 100 Hz with the subject maintaining the posture. This system has shown good accuracy and reliability for use in clinical practice and in research [21].

In the current study to assess the reliability of the MOCAP system, the standing posture was evaluated seven days later, at the same time and under identical conditions. All spherical reflective markers were attached to specific anatomic landmarks of the subjects by the same examiner.

No gap filling or filtering routines were performed on the data before they were output to an ASCII file for analysis. This ASCII file was loaded in Matlab ${}^{{\@setsize{\scriptsize}{8pt}{\viipt}{\@viipt}\textregistered}}$ software (The MathWorks, Natick, Massachusetts, USA) to calculate the thoracic and lumbar angles through the following equations:

Firstly, we calculated the direction of the vector $\overrightarrow{R1}=\overrightarrow{V2V1}:$

$\displaystyle\overrightarrow{V1}=v11\vec{i}+v12\vec{j}+v13\vec{k}$ $\displaystyle\overrightarrow{V2}=v21\vec{i}+v22\vec{j}+v23\vec{k}$ $\displaystyle\overrightarrow{r1}=\overrightarrow{V2}-\overrightarrow{V1}=(v21-% v11)\vec{i}+(v22-v12)\vec{j}$ $\displaystyle\qquad∼{}+(v23-v13)\vec{k}$ $\displaystyle\|\vec{r1}\|_{2}=$ $\displaystyle\sqrt{(v21\!-\!v11)^{2}+(v22\!-\!v12)^{2}+(v23\!-\!v13)^{2}}$ $\displaystyle\overrightarrow{R1}=\frac{\overrightarrow{r1}}{\|\overrightarrow{% r1}\|_{2}}=R11\vec{i}+R12\vec{j}+R13\vec{k}$

Then, we calculated the direction of the vector $\overrightarrow{R2}=\overrightarrow{V4V3}:$

$\displaystyle\overrightarrow{V3}=v31\vec{i}+v32\vec{j}+v33\vec{k}$ $\displaystyle\overrightarrow{V4}=v41\vec{i}+v42\vec{j}+v43\vec{k}$ $\displaystyle\overrightarrow{r2}=\overrightarrow{v4}-\overrightarrow{v3}=(v41-% v31)\vec{i}+(v42-v32)\vec{j}$ $\displaystyle\qquad∼{}+(v43-v33)\vec{k}$ $\displaystyle\|\overrightarrow{r2}\|_{2}=$ $\displaystyle\sqrt{(v41\!-\!v31)^{2}+(v42\!-\!v32)^{2}+(v43\!-\!v33)^{2}}$ $\displaystyle\overrightarrow{R2}=\frac{\overrightarrow{r2}}{\|\overrightarrow{% r2}\|_{2}}=R21\vec{i}+R22\vec{j}+R23\vec{k}$

Table 1

Test-retest reliability of MOCAP system for measuring spinal thoracic, lumbar and sacral inclination in standing posture*

	Thoracic spine	Lumbar spine	Sacral inclination
Session 1 (95% CI) degrees	40.76 $\pm$ 8.40 (24 to 61)	$-$ 33.36 $\pm$ 5.90 ( $-$ 22 to $-$ 47)	18.02 $\pm$ 3.92 (9 to 23)
Session 2 (95% CI) degrees	41.04 $\pm$ 8.70 (24 to 62)	$-$ 33.90 $\pm$ 5.49 ( $-$ 24 to $-$ 47)	18.08 $\pm$ 3.90 (9 to 23)
Systematic bias, degrees	$-$ 0.284	0.535	$-$ 0.054
Effect size $d$	0.023	0.067	0.010
Standard error, degrees	0.372	0.262	0.138
CV (%)	2.444	2.442	1.583
ICC (95% CI)	0.988 (0.974–0.995)	0.985 (0.963–0.994)	0.995 (0.986–0.998)

*Mean values $\pm$ standard deviation (s); CI $=$ confidence interval; CV (%) $=$ percentage coefficient of variation; ICC $=$ intraclass correlation coefficient.

Then, we can obtain the value of alpha1 (Thoracic curvature):

$\displaystyle\alpha_{1}={\rm acos}\left(\frac{\overrightarrow{R1}\odot% \overrightarrow{R2}}{\|\overrightarrow{R1}\|_{2}*\|\overrightarrow{R2}\|_{2}}\right)$ $\displaystyle={\rm acos}\!\left(\!\frac{R11\!*\!R21\!+\!R12\!*\!R22\!+\!R13\!*% \!R23}{\|\overrightarrow{R1}\|_{2}*\|\overrightarrow{R2}\|_{2}}\!\right)$ $\displaystyle\alpha_{1}={\rm abs}(\alpha_{1})$

The symbol $\odot$ es the scalar product and the abs ( ) operator is the absolute value.

Then, we calculated the direction of the vector $\overrightarrow{R3}=\overrightarrow{V6V5}:$

$\displaystyle\overrightarrow{V5}=v51\vec{i}+v52\vec{j}+v53\vec{k}$ $\displaystyle\overrightarrow{V6}=v61\vec{i}+v62\vec{j}+v63\vec{k}$ $\displaystyle\overrightarrow{r3}=\overrightarrow{V6}-\overrightarrow{V5}=(v61-% v51)\vec{i}+(v62-v52)\vec{j}$ $\displaystyle\qquad∼{}+(v63-v53)\vec{k}$ $\displaystyle\|\overrightarrow{r3}\|_{2}=$ $\displaystyle\sqrt{(v61\!-\!v51)^{2}+(v62\!-\!v52)^{2}+(v63\!-\!v53)^{2}}$ $\displaystyle\overrightarrow{R3}=\frac{\overrightarrow{r3}}{\|\overrightarrow{% r3}\|_{2}}=R31\vec{i}+R32\vec{j}+R33\vec{k}$

Then, we could obtain the value of alpha2 (lumbar curvature):

$\displaystyle\alpha_{2}={\rm acos}\left(\frac{\vec{R2}\odot\vec{R3}}{\|\vec{R2% }\|_{2}*\|\vec{R3}\|_{2}}\right)$ $\displaystyle={\rm acos}\!\left(\!\frac{R21\!*\!R31\!+\!R22\!*\!R32\!+\!R23\!*% \!R33}{\|\vec{R2}\|_{2}*\|\vec{R3}\|_{2}}\!\right)$ $\displaystyle\alpha_{2}={\rm abs}(\alpha_{2})$

We stablished the perpendicular vector to $\vec{i},\vec{k}/\vec{i}x\vec{k}=\vec{j}\equiv\bot$ as:

$\bot=0\vec{i}+1\vec{j}+0\vec{k}$

Then, we could obtain the value of alpha3 (sacral inclination):

$\displaystyle\alpha_{3}={\rm acos}\left(\frac{\overrightarrow{R3}\odot\bot}{\|% \overrightarrow{R3}\|_{2}*\|\bot\|_{2}}\right)$ $\displaystyle={\rm acos}\left(\frac{R31*0+R32*1+R33*0}{\|\overrightarrow{R3}\|% _{2}*\|\bot\|_{2}}\right)$ $\displaystyle\alpha_{3}={\rm abs}\left(\alpha_{3}\right)$

Table 2

Concurrent validity of MOCAP system for measuring spinal thoracic and lumbar curvatures and sacral inclination in standing posture*

		Thoracic spine	Lumbar spine	Sacral inclination
Gold standard	X Ray (95% CI), degrees	40.64 $\pm$ 7.93 (25 to 60)	$-$ 32.96 $\pm$ 5.28 ( $-$ 23 to $-$ 42)	18.48 $\pm$ 3.91 (9 to 23)
Session 1	Optitrack system (95% CI), degrees	40.76 $\pm$ 8.40 (24 to 61)	$-$ 33.36 $\pm$ 5.90 ( $-$ 22 to $-$ 47)	18.02 $\pm$ 3.92 (9 to 23)
	Systematic bias, degrees	$-$ 0.117	0.402	0.458
	Effect size $d$	0.010	0.050	0.072
	Standard error, degrees	0.515	0.643	0.179
	CV (%)	2.800	3.978	2.743
	ICC (95% CI)	0.975 (0.944–0.989)	0.912 (0.802–0.961)	0.988 (0.955–0.996)
Session 2	Optitrack system (95% CI), degrees	41.04 $\pm$ 8.70 (24 to 62)	$-$ 33.90 $\pm$ 5.49 ( $-$ 24 to $-$ 47)	18.08 $\pm$ 3.90 (9 to 23)
	Systematic bias, degrees	$-$ 0.401	0.937	0.404
	Effect size $d$	0.033	0.123	0.083
	Standard error, degrees	0.599	0.638	0.219
	CV (%)	3.080	4.28	3.282
	ICC (95% CI)	0.967 (0.926–0.986)	0.900 (0.775–0.956)	0.984 (0.954–0.994)

*Mean values $\pm$ standard deviation (s); CV (%) $=$ percentage coefficient of variation; CI $=$ confidence interval; ICC $=$ intraclass correlation coefficient.

The statistical analysis was performed using Statistical Package OS X, Version 22.0, Armonk, NY, USA). The data are shown as the mean(s) and standard deviation(s). Normality was analyzed using the Shapiro-Wilk test. All variables followed normal distributions. Paired Student’s $t$ -tests were used to detect any systematic differences (bias) between the test sessions (reliability) and tools (validity). The between-group effect sizes (Cohen’s $d$ ) were calculated using a pooled standard deviation. The effect sizes were evaluated as trivial (0–0.19), small (0.20–0.49), medium (0.50–0.79) and large ( $>$ 0.80) [22]. The intra-class correlation coefficients (ICCs) with 95% confidence intervals (CI) were used to examine the relative reliability of the Optitrack motion capture system. The absolute reliability was determined using the standard error of measurement [23] and the coefficients of variation that describe the intrasubject postural variation between sessions or instruments (expressed as a CV%) [24].

The concurrent validity of the Optitrack motion capture system was calculated using ICCs with a 95% CI. The statistical power and effect sizes were calculated using G $*$ Power 3.1 [25]. The level of significance was set at $P<$ 0.05. The statistical power was greater than 0.85 with the sample size of 25 participants.

3. Results

The test-retest mean values and reliability measures are shown in Table 1. Thoracic and lumbar spine and sacral inclination demonstrated excellent reliability, with mean ICCs levels greater than 0.980 and low CVs (mean: 2.15%). The systematic biases were not significant and very near 0, with a trivial effect size ( $<$ 0.07) and mean standard errors (mean: 0.257 ${}^{\circ}$ ).

Table 2 shows the mean values of the spinal thoracic and lumbar curvatures and of the sacral inclination evaluated by X ray (used as the gold standard) and motion capture with the Optitrack system in both sessions. The ANOVA analysis did not report any statistically significant differences in the comparisons, and the effect size was near 0. The systematic biases and mean random errors were lower than 1 ${}^{\circ}$ , with CVs lower than 5% and mean ICCs higher than 0.90 (0.775–0.994) for the thoracic and lumbar spine and sacral inclination between both techniques and sessions.

4. Discussion

In clinical practice and research, it is very important to evaluate sagittal spinal thoracic and lumbar curvature morphology because their physiological alterations have been associated with negative spinal consequences [3, 5]. Although radiographic methods are the “gold standard” for measurement of spinal posture and skeletal alignment [6, 7], their use is restricted because of the relative dose of radiation required and the necessary specific equipment.

The MOCAP system is increasingly being used in routine clinical practice and in research for gait and posture assessment because it requires less time and little equipment and is therefore less expensive and more accessible [21]. We do not know of any other studies examining the reliability and validity of spinal morphology evaluation with this MOCAP system. This study was designed to evaluate the test-retest reliability and validity of the MOCAP system to evaluate spinal sagittal thoracic and lumbar curvatures.

The major findings of this study were that the MOCAP system, using Optitrack infrared cameras, demonstrated excellent test-retest reliability and strong concurrent validity compared with a radiographic technique. The test-retest reliability (i.e., consistency or stability of the measurements) for thoracic and lumbar sagittal spinal curvatures and sacral inclination is critically important to ensure that observed differences in spinal morphology between testing sessions are not because of systematic bias, such as a learning effect or fatigue, or a standard error of measurement due to possible biological or mechanical variations [24]. For both sagittal spinal curvatures (thoracic and lumbar) and sacral inclination, we found that the systematic bias and SEM was lower than 1 ${}^{\circ}$ between the testing sessions. The test-retest CVs of thoracic and lumbar curvatures and sacral inclination were lower than 2.5%. Atkinson and Nevill [24] reported that an analytical goal of CV $\leqslant$ 10% might be an indicator of acceptable agreement. It has been suggested that ICC values above 0.75 could be considered reliable, and this index should be at least 0.90 for most clinical applications [26]. In our study results showed ICC values higher than 0.98 for thoracic and lumbar curvatures and sacral inclination. These values are higher than those of previous studies that evaluated a new system for measuring spinal curvatures [15, 27]. The high ICCs, low CVs, absence of any significant difference and similar values between the trials indicated that the MOCAP system data were reliable.

The concurrent validity of the MOCAP system was very high for both sessions. We found the following results: a less than 1 ${}^{\circ}$ systematic bias, $P>$ 0.05, with a standard error $<$ 1 ${}^{\circ}$ ; CV $<$ 5%; and ICC greater than 0.9 in both sessions with respect to the radiographic technique (used as the gold standard).

In healthy individuals, we found mean values of approximately 40 ${}^{\circ}$ and 33 ${}^{\circ}$ for thoracic kyphosis and lumbar lordosis, respectively, and 18 ${}^{\circ}$ for pelvis inclination. These results are in concordance with the range of spinal values for adults reported by previous studies [1, 2, 15, 20, 27]. With these results, we report that the MOCAP system, used with our algorithms, is a valid system for evaluating spinal curvatures and sacral inclination.

Regarding the evaluated posture, when taking a lateral whole-spine radiograph, the best body position is a nearly relaxed standing position to gain a more ecologic visualization of the neutral spine. The most appropriate arm position during radiography for finding adequate lateral radiographic visualization of the spine has not been determined. Suzuki et al. [28] found that a clasped position could be used effectively and reliably for measurement of sagittal alignment for the lumbar region in adults. They did not evaluate the thoracic spine. In this study, the posture adopted for the radiographs was a relaxed standing posture with both arms forward and 90 ${}^{\circ}$ of shoulder flexion. This posture has been used in the literature [29, 30]. Zaina et al. [30] reported that arm position changes the spinal posture because an optimal position that is comparable to normal standing does not exist. For that reason, these authors affirm that a surface device might be a more ecologic measurement instrument than radiographs because they allow the patient to maintain a normal position of their arms, resulting in greater reliability. We considered that using a support for relaxing the arms with shoulders flexed to 90 ${}^{\circ}$ could be more reproducible for this reliability and validity study.

In line with the results obtained in this study, the MOCAP system delivered consistently reliable and valid results for standing curvatures in comparison with a radiographic technique. The values obtained are in agreement with those previously reported in the literature. These results suggest that this system could be used with confidence in research and clinical environments for the measurement of the morphology in the sagittal spinal curvatures but not as a diagnostic system. This system has the advantage of being able to be used to evaluate spinal curvatures in movement, reporting a complete understanding of the spinal curvature morphology in functional movements because it is a non-invasive tool.

Conflict of interest

None to report.

Funding

Proyectos I $+$ D $+$ I Ministerio de Economía y Competitividad. Gobierno de España. Referencia: DEP 2016-80296-R (AEI/FEDER, UE).

Footnotes

Acknowledgments

The authors thank the Vithas-Virgen del Mar Hospital for contributing the radiological equipment for this study.

References

Mejia

Hennrikus

Schwend

Emans

. A prospective evaluation of idiopathic left thoracic scoliosis with magnetic resonance imaging. Journal of Pediatric Orthopaedics. 1996; 16(3): 354-358.

Tüzün

Yorulmaz

Cindaş

Vatan

. Low back pain and posture. Clinical Rheumatology. 1999; 18(4): 308-312.

Wilke

Neef

Caimi

Hoogland

Claes

. New in vivo measurements of pressures in the intervertebral disc in daily life. Spine. 1999; 24(8): 755-762.

Solomonow

Zhou

Baratta

Burger

. Biomechanics and electromyography of a cumulative lumbar disorder: Response to static flexion. Clinical Biomechanics. 2003; 18(10): 890-898.

Beach

Parkinson

Stothart

Callaghan

. Effects of prolonged sitting on the passive flexion stiffness of the in vivo lumbar spine. Spine Journal. 2005; 5(2): 145-154.

Harrison

Cailliet

Harrison

Janik

Holland

. Reliability of centroid, Cobb, and Harrison posterior tangent methods: which to choose for analysis of thoracic kyphosis. Spine. 2001; 26(11): E227-E234.

Harrison

Cailliet

Janik

Holland

. Radiographic analysis of lumbar lordosis: Centroid, Cobb, TRALL, and Harrison posterior tangent methods. Spine. 2001; 26(11): E235-E242.

Korovessis

Petsinis

Papazisis

Baikousis

. Prediction of thoracic kyphosis using the Debrunner kyphometer. Journal of Spinal Disorders. 2001; 14(1): 67-72.

D’Osualdo

Schierano

Iannis

. Validation of clinical measurement of kyphosis with a simple instrument, the arcometer. Spine. 1997; 22(4): 408-413.

10.

Walicka-Cupryś

Drzał-Grabiec

Mrozkowiak

. Comparison of parameters of spinal curves in the sagittal plane measured by photogrammetry and inclinometry. Ortopedia Traumatologia Rehabilitacja. 2013; 15(5): 429-437.

11.

Sedrez

Candotti

Rosa

Medeiros

Marques

Loss

. Test-retest, inter- and intra-rater reliability of the flexicurve for evaluation of the spine in children. Brazilian Journal of Physical Therapy. 2016; 20(2): 142-147.

12.

Adam

Izatt

Harvey

Askin

. Variability in Cobb angle measurements using reformatted computerized tomography scans. Spine. 2005; 30(14): 1664-1669.

13.

Barczyk-Pawelec

Piechura

Dziubek

Rożek

. Evaluation of isokinetic trunk muscle strength in adolescents with normal and abnormal postures. Journal of Manipulative and Physiological Therapeutics. 2015; 38(7): 484-492.

14.

Burnett

Cornelius

Dankaerts

O’Sullivan

. Spinal kinematics and trunk muscle activity in cyclists: A comparison between healthy controls and non-specific chronic low back pain subjects – a pilot investigation. Manual Therapy. 2004; 9(4): 211-219.

15.

Kellis

Adamou

Tzilios

Emmanouilidou

. Reliability of spinal range of motion in healthy boys using a skin-surface device. Journal of Manipulative and Physiological Therapeutics. 2008; 31(8): 570-576.

16.

Sheeran

Sparkes

Busse

van Deursen

. Preliminary study: Reliability of the spinal wheel. A Novel Device to Measure Spinal Postures Applied to Sitting and Standing. 2010; 19(6): 995-1003.

17.

Dunk

Lalonde

Callaghan

. Implications for the use of postural analysis as a clinical diagnostic tool: Reliability of quantifying upright standing spinal postures from photographic images. Journal of Manipulative and Physiological Therapeutics. 2005; 28(6): 386-392.

18.

Barrett

McCreesh

Lewis

. Reliability and validity of non-radiographic methods of thoracic kyphosis measurement: A systematic review. Manual Therapy. 2014; 19(1): 10-17.

19.

Bolink

Naisas

Senden

Essers

Heyligers

Meijer

Grimm

. Validity of an inertial measurement unit to assess pelvic orientation angles during gait, sit-stand transfers and step-up transfers: Comparison with an optoelectronic motion capture system. Medical Engineering & Physics. 2016; 38(3): 225-231.

20.

Keller

Colloca

Harrison

Janik

. Influence of spine morphology on intervertebral disc load and stresses in asymptomatic adults: Implications for the ideal spine. Spine J. 2005; 5(3): 297-309.

21.

Carse

Meadows

Bowers

Rowe

. Affordable clinical gait analysis: An assessment of the marker tracking accuracy of a new low-cost optical 3D motion analysis system. Physiotherapy. 2013; 99(4): 347-351.

22.

Cohen

. Statistical power analysis for the behavioral science. 2nd Hillsdale, NJ: Lawrence Erlbaum Associates; 1988.

23.

van de Ven-Stevens

Munneke

Terwee

Spauwen

van der Linde

. Clinimetric properties of instruments to assess activities in patients with hand injury: A systematic review of the literature. Archives of Physical Medicine and Rehabilitation. 2009; 90(1): 151-169.

24.

Atkinson

Nevill

. Statistical methods for assessing measurement error (reliability) in variables relevant to sports medicine. Sports Medicine. 1998; 26(4): 217-238.

25.

Faul

Erdfelder

Lang

Buchner

. G*Power 3: A flexible statistical power analysis program for the social, behavioral, and biomedical sciences. Behaviour Research Methods. 2007; 39(2): 175.

26.

Walmsley

Amell

. The application and interpretation of intraclass correlations in the assessment of reliability in isokinetic dynamometry. Isokinetics and Exercise Science. 1996; 6(2): 117-124.

27.

Mannion

Knecht

Balaban

Dvorak

Grob

. A new skin-surface device for measuring the curvature and global and segmental ranges of motion of the spine: Reliability of measurements and comparison with data reviewed from the literature. European Spine Journal. 2004; 13(2): 122-136.

28.

Suzuki

Endo

Mizuochi

Kobayashi

Tanaka

Yamamoto

. Clasped position for measurement of sagittal spinal alignment. European Spine Journal. 2010; 19(5): 782-786.

29.

Faro

Marks

Pawelek

Newton

. Evaluation of a functional position for lateral radiograph acquisition in adolescent idiopathic scoliosis. Spine. 2004; 29(20): 2284-2289.

30.

Zaina

Pizzetti

Donzelli

Negrini

. Why X-rays are not reliable to assess sagittal profile: A cross sectional study. Studies in Health Technology and Informatics. 2012; 176: 268-272.