Validity and Reliability of the 3D Motion Analyzer in Comparison

with the Vicon Device for Biomechanical Pedalling Analysis

Anthony Bouillod

1,2,3

, Antony Costes

, Georges Soto-Romero

3,5

, Emmanuel Brunet

and Frederic Grappe

1,6

EA4660, C3S Health - Sport Department, Sports University, Besancon, France

French Cycling Federation, Saint Quentin en Yvelines, France

LAAS-CNRS, Université de Toulouse, CNRS, Toulouse, France

Université de Toulouse, UPS, PRISSMH, Toulouse, France

ISIFC - Génie Biomédical, 23 Rue Alain Savary, Besançon, France

Professional Cycling Team FDJ, Moussy le Vieux, France

Keywords: Bike Fitting, 3D Kinematics, Comparison, Laboratory, Cycling.

Abstract: The present work aimed to assess the validity and reliability of the 3D motion analyzer (Shimano Dynamics

Lab, Sittard, Netherland) during laboratory cycling tests in comparison with the Vicon device (Vicon

Motion Systems Ltd. Oxford, UK). Three cyclists were required to complete one laboratory cycling test at

three different pedalling cadence and at a constant power output. Kinematic measurements were collected

simultaneously from 3D motion analyzer and Vicon devices and performed five times for each pedalling

cadence. The two systems showed a high reliability with excellent intraclass correlation coefficients for

most kinematic variables. Moreover, this system was considered as valid by considering the error due to the

initial markers placement. Experts and scientists should use the Vicon system for the purpose of research

whereas the 3D motion analyzer could be used for bike fitting.

1 INTRODUCTION

Bike fitting is an important process to adjust the

geometry of the bike and its components to the

needs of the cyclist. Optimal position on the bicycle

may be considered as a position in which force

application and comfort are maximised, whilst

resistive forces and risk of injury are minimised, in

order to maximise bicycle velocity (Iriberri et al.,

2008). The manipulation of a single variable such as

saddle height can improve performance within

cycling economy (Peveler and Green, 2010) and

power output in anaerobic exercises (Peveler et al.,

2007).

Numerous methodologies and systems have

been proposed to perform bike fitting (Holmes et al.,

1994, Iriberri et al., 2008, Nordeen-Snyder, 1977).

However, different kinematic systems do not

necessary provide the same results (Fonda et al.,

2014). Umberger and Martin (Umberger and Martin,

2001) reported that no significant difference exist

between 3D and 2D kinematic systems whereas

Fonda et al. (Fonda et al., 2014) measured

significant differences between the two systems. The

3D motion analyzer (Shimano Dynamics Lab,

Sittard, Netherland) is a new kinematic system

positioned in the sagittal plane and tracking LED

markers attached to the skin.

This study aimed to assess the validity and

reliability of the 3D motion analyzer in comparison

with the Vicon device for biomechanical pedalling

analysis. We hypothesized that 1) the kinematic

variables measured by the two systems would be

similar and 2) the two systems will achieve an

excellent reliability.

2 METHODS

Three cyclists volunteered to participate in the study.

Prior to testing and after having received a full

explanation of the nature and purpose of the study,

the participants gave their written informed

consents. The participants performed one testing

session with the same road-racing bicycle (Lapierre

Pulsium, Dijon, France). The validity and reliability

Bouillod, A., Costes, A., Soto-Romero, G., Brunet, E. and Grappe, F.

Validity and Reliability of the 3D Motion Analyzer in Comparison with the Vicon Device for Biomechanical Pedalling Analysis.

DOI: 10.5220/0006088200630066

In Proceedings of the 4th International Congress on Sport Sciences Research and Technology Support (icSPORTS 2016), pages 63-66

ISBN: 978-989-758-205-9

of the 3D motion analyzer was investigated on an

Elite Novo Force home-trainer (Elite, Fontaniva,

Italy) (figure 1) at three different pedalling cadences

(60, 90 and 120 rpm) and at a constant power output

(200 W) measured by a SRM power meter (SRM,

Schoberer Rad Messtechnich, Julich, Germany).

Kinematic measurements were performed five times

for each pedalling cadence resulting in 15 different

data sets by participant, each data set lasting 10

seconds.

Kinematic analysis of the cyclists’ right side

was performed assuming symmetry of motion

between left and right sides (Heil et al., 1997,

Garcia-Lopez et al., 2015) and using the 3D motion

analyzer. Height LED markers with built-in probe

were attached to the skin of the cyclists (fifth

metatarsal head, calcaneus, lateral malleolus, lateral

femoral epicondyle, greater trochanter, acromion,

lateral epicondyle of the humerus and styloid

process of the ulna) (Bini et al., 2010a, Ferrer-Roca

et al., 2012). The 3D sensor was positioned 2 m

away from the sagittal plane and was calibrated

before the study as recommended by the

manufacturer. Automatic tracking, processing and

analysing data were performed by a specific

software (Bikefitting.com, Version 2.1.5, Shimano

Dynamics Lab, Sittard, Netherland).

Kinematic data were also collected from 12

passive markers recorded by twelve infrared

cameras (Vicon Motion Systems Ltd. Oxford, UK).

These markers were attached in the same line,

interspersed within active ones (LED markers). The

Vicon system (using Nexus 1.7.1 software) is a

marker-based motion capture system acknowledged

as a reference. This motion capture system carries 12

MX3+ cameras with a frequency of 200 Hz, a

millimeter accuracy and a resolution of 659 × 494

pixels each. The data processing was performed

using custom-made code written in Matlab software

(Matlab Release 2014a, The MathWorks, Inc.,

Natick, Massachusetts, USA).

Shoulder, forearm, elbow, torso, hip, knee,

ankle and foot (vertical and lateral) angles were

determined. Angular position values of the hip, knee

and ankle were expressed as flexion (minimum

angle) and extension (maximum angle). Knee lateral

travel and knee travel tilt were also measured during

the study.

3 RESULTS AND DISCUSSION

The two devices showed a high reliability with no

significant difference and excellent intraclass

correlation coefficients (Cicchetti, 1994) for most

kinematic variables (table 1). This results confirmed

our hypothesis showing that the two systems

achieved a high reliability.

Table 1: Intraclass correlation coefficients (ICC) for the

two systems and for all kinematic variables.

Variable

ICC

Vicon

Motion

Analyzer

Shoulder angle (°) 0.90 0.93

Forearm angle (°) 0.35 0.63

Elbow angle (°) 0.74 0.59

Torso angle (°) 0.79 0.95

Hip angle extension (°) 0.86 0.63

Hip angle flexion (°) 0.94 0.97

Knee angle extension (°) 0.96 0.98

Knee angle flexion (°) 0.88 0.97

Knee lateral travel (mm) 0.84 0.88

Knee travel tilt (°) 0.99 0.92

Foot vertical angle (°) 0.89 0.66

Foot lateral angle (°) 0.97 0.94

Ankle angle maximum (°) 0.97 0.92

Ankle angle minimum (°) 0.84 0.89

Ankle range (°) 0.85 0.72

Figure 1: settings for the experiment.

All kinematic variables collected with the 3D

motion analyzer were significantly correlated with

those collected with the Vicon device (table 2)

except for knee angle flexion and foot lateral angle.

However, some statistical differences have been

reported suggesting that the 3D motion analyzer

measurements were significantly different that those

measured with the Vicon system. The present study

icSPORTS 2016 - 4th International Congress on Sport Sciences Research and Technology Support

Table 2: Comparative statistics of kinematic variables (n = 3) measured during the session between the Vicon system and

the 3D motion analyzer system for all pedalling cadences. Values are reported as mean ± SD.

Variable Vicon

Motion

Analyzer

Statistical

difference

Linear

regression

)

Bias 95% CI- 95% CI+

Shoulder angle (°) 77.7 ± 4.5 74.9 ± 4.0 ** 0.96** -2.8 -4.7 -0.9

Forearm angle (°) 41.3 ± 2.1 39.7 ± 2.7 ** 0.55** -1.6 -5.1 1.9

Elbow angle (°) 158.4 ± 4.8 157.2 ± 3.4 * 0.66** -1.2 -6.9 4.4

Torso angle (°) 45.5 ± 2.5 45.2 ± 3.3 n.s. 0.86** -0.3 -3.0 2.4

Hip angle extension (°) 105.7 ± 4.7 104.6 ± 3.8 ** 0.70** -1.0 -6.1 4.0

Hip angle flexion (°) 59.1 ± 3.4 62.3 ± 1.2 n.s. 0.18* 3.2 -3.0 9.3

Knee angle extension (°) 40.2 ± 2.6 46.6 ± 5.0 ** 0.76** 6.3 0.5 12.2

Knee angle flexion (°) 119.5 ± 2.1 114.7 ± 3.3 ** 0.01 -4.8 -12.4 2.8

Knee lateral travel (mm) 18.3 ± 6.5 22.2 ± 9.2 ** 0.66** 3.9 -6.7 14.5

Knee travel tilt (°) 3.8 ± 1.9 3.6 ± 2.5 n.s. 0.23** -0.2 -7.8 7.3

Foot vertical angle (°) 26.6 ± 2.1 25.9 ± 2.9 * 0.39** -0.7 -5.1 3.7

Foot lateral angle (°) 5.7 ± 1.8 8.3 ± 2.0 ** 0.05 2.6 -2.0 7.2

Ankle angle maximum

(°)

101.5 ± 6.3 99.3 ± 5.7 ** 0.92** -2.1 -5.7 1.4

Ankle angle minimum (°) 75.4 ± 3.5 75.6 ± 4.0 n.s. 0.87** 0.2 -2.6 3.0

Ankle range (°) 26.0 ± 4.2 23.6 ± 3.6 ** 0.92** -2.4 -4.9 0.1

* Significant at p < 0.05; ** Significant at p < 0.001; 95% CI = 95% Confidence Interval.

confirmed that various motion capturing systems do

not necessarily provide the same results as they

work on different basis (Fonda et al., 2014). This is

of practical importance when adjusting body

position. Even though most of the kinematic

variables were significantly different, these

differences are often less than 3°.

These results underlined the importance of the

markers placement for comparative and statistical

analysis between the two systems. Considering the

error due to the initial markers placement (obtained

for each cyclist) as an offset, we could compensate

the significant difference obtained for knee angle

extension (6.3°). To a lesser extent, the differences

measured between the two systems could be

influenced by some movements of the 3D motion

analyzer markers (altering the alignment with the

Vicon markers) during dynamic measurements. Our

results indicated that with a cycling specific motion

analysis tool and easy post-processing analysis, we

are able to obtain reliable and useful data for bike

fitting, in comparison with a full 3D motion capture

system.

There was a significant increase in knee

extension (from 38.4 to 42.3°) and knee flexion

(from 118.5 to 120.6°) mean angles with increasing

pedalling cadence only with the Vicon device.

Additionally, foot vertical and ankle range mean

angles were significantly increased with pedalling

cadence for both the 3D motion analyzer and the

Vicon system. Divergence among studies has been

observed regarding the contribution of each joint

(Hoshikawa et al., 2007, Mornieux et al., 2007, Bini

et al., 2010b). The current study indicated that knee

joint changed with increasing pedalling cadence

whereas several authors (Bini et al., 2010b, Ericson,

1988) have reported no change. However, ankle

range increased when pedalling cadence was

increased from 60 to 120 rpm. This result is in

accordance with previous studies (Ericson, 1988,

Hoshikawa et al., 2007, Mornieux et al., 2007, Bini

et al., 2010b) suggesting that ankle joint muscles

control the pedal force application.

Note that this study is limited to only three

participants and is therefore considered as

preliminary. Nevertheless, the study design

provided a large number of measurement over a

variety of pedalling cadences typically generated by

elite athletes.

Validity and Reliability of the 3D Motion Analyzer in Comparison with the Vicon Device for Biomechanical Pedalling Analysis

4 CONCLUSIONS

The 3D motion analyzer showed a high reliability.

Moreover, this system was considered as valid after

compensation of the operator dependent error due to

the initial markers alignment between the two

systems.

Experts and scientists should use the Vicon

system for the purpose of research whereas the 3D

motion analyzer could be used for bike fitting. Bike

fitting experts could employ a correction factor for

each kinematic variables using the constant bias

measured in our study. Additionally, these experts

must standardize the pedalling cadence during bike

fitting sessions considering that the contribution of

the ankle joint was influenced by the pedalling

cadence.

REFERENCES

bini, R. R., Diefenthaeler, F. & Mota, C. B. 2010a.

Fatigue effects on the coordinative pattern during

cycling: kinetics and kinematics evaluation. J

Electromyogr Kinesiol, 20, 102-7.

Bini, R. R., Tamborindeguy, A. C. & Mota, C. B. 2010b.

Effects of saddle height, pedaling cadence, and

workload on joint kinetics and kinematics during

cycling. J Sport Rehabil, 19, 301-14.

Cicchetti, D. V. 1994. Guidelines, criteria, and rules of

thumb for evaluating normed and standardized

assessment instruments in psychology. Psychological

assessment, 6, 284.

Ericson, M. O. 1988. Mechanical muscular power output

and work during ergometer cycling at different work

loads and speeds. European journal of applied

physiology and occupational physiology, 57, 382-387.

Ferrer-Roca, V., Roig, A., Galilea, P. & Garcia-Lopez, J.

2012. Influence of saddle height on lower limb

kinematics in well-trained cyclists: static vs. dynamic

evaluation in bike fitting. J Strength Cond Res, 26,

3025-9.

Fonda, B., Sarabon, N. & Li, F. X. 2014. Validity and

reliability of different kinematics methods used for

bike fitting. J Sports Sci, 32, 940-6.

Garcia-Lopez, J., Diez-Leal, S., Ogueta-Alday, A.,

Larrazabal, J. & Rodriguez-Marroyo, J. A. 2015.

Differences in pedalling technique between road

cyclists of different competitive levels. J Sports Sci,

1-8.

Heil, D. P., Derrick, T. R. & Whittlesey, S. 1997. The

relationship between preferred and optimal

positioning during submaximal cycle ergometry. Eur

J Appl Physiol, 75, 160-165.

Holmes, J. C., Pruitt, A. L. & Whalen, N. J. 1994. Lower

extremity overuse in bicycling. Clin Sports Med, 13,

187-205.

Hoshikawa, H., Takahashi, K., Ohashi, K. & Tamaki, K.

2007. Contribution of the ankle, knee, and hip joints

to mechanical energy in cycling. Journal of

biomechanics, 40, S750.

Iriberri, J., Muriel, X. & Larrazabal, I. 2008. The bike fit

of the road professional cyclist related to

anthropometric measurements and the torque of de

crank. The Engineering of Sport, 7, 483-488.

Mornieux, G., Guenette, J. A., Sheel, A. W. & Sanderson,

D. J. 2007. Influence of cadence, power output and

hypoxia on the joint moment distribution during

cycling. Eur J Appl Physiol, 102, 11-8.

Nordeen-Snyder, K. S. 1977. The effect of bicycle seat

height variation upon oxygen consumption and lower

limb kinematics. Medicine and Science in Sports, 9,

113-117.

Peveler, W. W. & Green, M. 2010. The effect of extrinsic

factors on simulated 20-km time trial performance.

The Journal of Strength & Conditioning Research,

24, 3265-3269.

Peveler, W. W., Pounders, J. D. & Bishop, P. A. 2007.

Effects of saddle height on anaerobic power produ-

ction in cycling. J Strength Cond Res, 21, 1023-7.

Umberger, B. R. & Martin, P. E. 2001. Testing the planar

assumption during ergometer cycling. Journal of

Applied Biomechanics, 17, 55-62.

icSPORTS 2016 - 4th International Congress on Sport Sciences Research and Technology Support