Program & abstracts

Document technical information

Format pdf
Size 3.5 MB
First found May 22, 2018

Document content analysis

Category Also themed
Language
English
Type
not defined
Concepts
no text concepts found

Persons

Johann Strauss II
Johann Strauss II

wikipedia, lookup

Andrei Konchalovsky
Andrei Konchalovsky

wikipedia, lookup

Organizations

Places

Transcript

FACULTY OF HEALTH AND MEDICAL SCIENCES
UNIVERSITY OF COPENHAGEN
Program & abstracts
5 th Annual Meeting of the Danish Society of Biomechanics
25 October 2013
Haderup Auditorium
The Panum Institute
University of Copenhagen
U NI V E R S I T Y
O F
CO P E NH A G E N
Welcome
Dear Colleagues,
It is a great pleasure to welcome you to the 5th annual meeting of the Danish Society of
Biomechanics (DBS). The society was founded at the first annual meeting in 2009 at the Parker
Institute, Frederiksberg Hospital, Copenhagen and the following meetings have been touring round
Denmark visiting Aalborg, Odense, and Aarhus. Now we are back in Copenhagen – this time at the
Panum Institute, University of Copenhagen.
As usual the annual meeting will consist of a Keynote Lecture, a Steno Lecture, poster sessions,
podium presentations and the general assembly for members of DBS. This time the oral
presentations will focus on “what is going on in the field of biomechanics in Denmark and in
countries where Danes are doing biomechanics”. We are grateful that all the speakers accepted to
give us a brief presentation of their biomechanical environment and the projects they are working
on. We are also honoured to welcome our keynote speaker Dr. Daniel Benoit, University of Ottawa,
Canada and our Steno lecturer Dr. Troels Kardel, Copenhagen.
We would also like to thank you for your contributions to this meeting and we look greatly forward
to an inspiring day together with you.
Finally, special thanks to our sponsor Velamed.
Best wishes,
The organising committee:
Henrik Koblauch
Peter Raffalt
Martin S. Christiansen
Erik B. Simonsen
Tine Alkjær (DBS board member)
U NI V E R S I T Y
O F
CO P E NH A G E N
Practical Information
Location
The Main Entrance is on Nørre Allé 20. Please follow the red line on the map to the Haderup
Auditorium.
Parking is very limited and quiet costly (~15 kr/hour), so we encourage you not to use your own car
as means of transport.
There are many bus connections to Frederik Bajers Plads (Nørre Allé – Tagensvej intersection).
From there it is a two minute walk to the main entrance.
Internet
You will receive a personal code to the wireless internet (access point: KU Guest). Your username
is the email you signed up with and the password is individually supplied (you will receive an email
with a code).
U NI V E R S I T Y
O F
CO P E NH A G E N
Keynote Lecture
Dr. Daniel is this year’s keynote speaker at the DBS annual meeting. Daniel Benoit is Associate
Professor from School of Rehabilitation Sciences, University of Ottawa. Dr. Daniel Benoit received
his B.Sc. in Human Kinetics from the University of Ottawa and his M.Sc. in Human Biodynamics
(Biomechanics) from McMaster University. He pursued his career in Europe as director of a clinical
biomechanics laboratory in Perugia-Italy where, in addition to clinical responsibilities, he
investigated in vivo ACL strain and knee joint loading. Dr. Benoit is currently on sabbatical and has
joined the biomechanical division at the Department of Neuroscience and Pharmacology,
University of Copenhagen, as a visiting professor where he collaborates with Erik B. Simonsen,
Tine Alkjær and Michael Krogsgaard (Bispebjerg Hospital).
New insights on the role of muscles in knee stabilisation
The knee joint is the most researched joint in the human body. Its role is essential for mobility and
injuries to this joint therefore can have a significant effect on lifestyle and independence. The most
commonly studied injury to the knee is the anterior cruciate ligament (ACL) injury, which leads to
joint instability and in most cases knee osteoarthritis (OA). The ACL injury and knee OA are more
prevalent in females and, once again, the knee is less stable in this population. Muscles are the
only active modulator of joint stability and thus neuromuscular control can be considered the only
path which might allow an intervention to modify this stability as a function of environmental
requirements for both prophylactic and rehabilitation purposes. To do so, a thorough understanding
of the functional role of the muscles surrounding the knee is required. In this presentation I will
explore the literature surrounding this issue and in particular my past experiences on the topic.
U NI V E R S I T Y
O F
CO P E NH A G E N
Steno Lecture
Dr. Troels Kardel is the recipient of the Steno award 2013. Troels Kardel has been interested in the
research of Steno and is the editor of a recent biography of Niels Steensen (Steno). At the DBS
web site (http://www.danskbiomekaniskselskab.dk/) you can find more information about Troels
Kardel’s research about Steno. Congratulations on the Steno award.
On Steno and his myology. An overview of 40 years’ research by Troels Kardel
The 16th -17th centuries are considered the period of Early Modern Science in which period a
transition took place from scholastic discussions of natural phenomena founded on ancient authors
and the Scriptures to inquiries based on observation excluding immanent powers and inherent
qualities. In astronomy and physics the forerunners were Kopernicus, Galileo and Newton, and in
medicine Vesalius and Harvey. There is no sharp line drawn - new ideas such as those expressed
by Bacon and Descartes worked alongside traditional views.
It took time and pain before the heliocentric system became generally accepted. It is less known
that Harvey met widespread opposition, and that his concept on the circulation of the blood, which
is not an observation but an observation supported theory, was generally accepted only after his
lifetime. It is practically unknown that a theory of muscle contraction based on fibre shortening and
the macrostructure of muscle was sent into limbo for no less than three hundred years before full
acceptance, a theory that was rejected by eminent scientists even without any interference by
clerics.
The exponent of the new ideas on muscles, the Musculi systema novum, is Niels Stensen, called
Steno (1638-1686). The son of a goldsmith in Copenhagen he studied medicine in his home-town
during a period of maximal unrest. The city was under siege by the Swedish king. He studied
anatomy in Leiden under the famous anatomist and iatrochemist Sylvius. He went to Paris where
he gave a still readable discourse on the anatomy of the brain. He then went on to become
scientist at the court of the Grand Duke of Tuscany Ferdinando II who in his youth had supported
Galileo’s work. During busy research in Tuscany on muscles, on geology and on female
reproductive organs Steno converted to the Catholic church. He became later bishop in Germany
where he died in Schwerin aged 48. His body was transferred to Florence at the expense of the
Grand Duke. Niels Stensen was beatified in the Catholic church in 1988.
U NI V E R S I T Y
O F
CO P E NH A G E N
5 th annual meeting of the Danish Society of Biomechanics
Program
09.30-10.00
Welcome and registration (Tine Alkjær)
10.00-10.40
Keynote lecture: Dr. Daniel Benoit, (University of Ottawa, Canada): New insights on
the role of muscles in knee stabilisation. (Chair: Tine Alkjær)
10.40-11.30
Poster session (Odd poster numbers) & coffee
11.30-12.10
Steno lecture: Dr. Troels Kardel (Copenhagen): On Steno and his Myology – Finds
and Essentials in 40 Years’ research. (Chair: Finn Bojsen-Møller)
12.10-13.00
Lunch & Poster session (Even poster numbers)
13.00-13.40
General Assembly meeting DBS (only for DBS members)
13.40-14.40
Oral presentations, session I. (Chair: Erik B. Simonsen)
13.40-14.00
Biomechanics in Aalborg: Interaction between hardcore computational
biomechanics and complicated experiments, Mark de Zee, Aalborg University.
14.00-14.20
Biomechanical analyses of Orthopaedic Research Unit, Odense University Hospital
– current methods and ongoing projects, Anders Holsgaard Larsen, The
Orthopaedic Research Unit, Odense University Hospital.
14.20-14.40
The Gait Analysis Laboratory, Copenhagen University Hospital at Hvidovre research methods & projects, Hanne Bloch, Hvidovre Hospital.
14.40-15.00
Coffee break
15.00-16.00
Oral presentations, session II. (Chair: Maj Halling-Thomsen).
15.00-15.20
Biomechanical Laboratory at Norwegian School of Sport Sciences, Jens BojsenMøller, Norwegian School of Sport Sciences, Norway.
15.20-15.40
Understanding Human Movement Behavior - current and future research, Peter
Raffalt, Department of Neuroscience and Pharmacology, University of Copenhagen.
15.40-16.00
Application of cluster based exposure variation analysis in biomechanics and sport
science, Afshin Samani, Center for Sensory-Motor Interaction (SMI), Dept. of Health
Science and Technology, Aalborg University.
16.00-16.10
Closing remarks (Tine Alkjær)
16.10-?
Beer, snacks, and small talk
U NI V E R S I T Y
O F
CO P E NH A G E N
List of Abstracts
Oral
Application of cluster based exposure variation analysis in biomechanics and sport science
Afshin Samani
Posters
Session I = Odd numbers, Session II = Even numbers
Board 1
Motor control during walking in patients with knee osteoarthritis
Helle Dalsgaard, Tine Alkjær, Erik B. Simonsen, Marius Henriksen
Board 2
A comparison of accuracy and precision of 5 gait-event detection algorithms
from motion capture in horses during over ground trot
Jenny Katrine Boyea, Emil Olsen, Thilo Pfau, Maj Halling Thomsen
Board 3
Weight bearing knee joint stabilisation strategies differ between younger,
older, and osteoarthritic populations
Teresa E. Flaxman, Andrew J.J. Smith, Daniel L. Benoit
Board 4
FE model of the hip based on clinically recorded CT
Tina Skytte
Board 5
Does quadriceps-strengthening exercise affect quadriceps force, power and
work during stair ascent in adults with knee osteoarthritis?
Josh Leonardis, Patrick Rider, John Norbury, Jens Aaboe, Robin Christensen,
Henning Bliddal, Marius Henriksen, Paul DeVita
Board 6
Application of ergovar, an interactive open-source toolbox in matlab, in
biomechanics
Afshin Samani
Board 7
Non-negative matrix factorization algorithm as a way to analyze muscle
synergies: Application to bench press exercise
Mathias Kristiansen, Pascal Madeleine, Ernst Albin Hansen, Afshin Samani
Board 8
Ankle bracing changes lower extremity alignment during landing
Ilias Theodorakos, Jan Rueterbories, Michael S. Andersen, Mark de Zee, Uwe G.
Kersting
Board 9
Using isokinetic dynamometer measurements to determine muscle strength
characteristics of elite cyclists: a preliminary study
Frederik Heinen, Carsten Møller Mølgaard, John Rasmussen, Mark de Zee
Board 10
Lumbar load in common work tasks for airport baggage handlers – a 3D
dynamic analysis Henrik Koblauch, Stine H Bern, Charlotte Brauer, Sigurd
U NI V E R S I T Y
O F
CO P E NH A G E N
Mikkelsen, Mark de Zee, Lau C Thygesen, Karin Helweg-Larsen, Tine Alkjær,
Michael S Andersen, Erik B Simonsen
Board 11
High intensity eccentric training in the neck-shoulder region
Steffen Vangsgaard, Ernst Albin Hansen, Pascal Madeleine
Board 12
Sex differences in muscular load among house painters performing identical
work tasks
Thomas Heilskov-Hansen, Jacob Meyland, Tine Alkjær, Henrik Koblauch, Sigurd
Mikkelsen, Susanne Wulff Svendsen, Jane Frølund Thomsen, Gert-Åke Hansson,
Erik Bruun Simonsen
Board 13
Heavy strength training and voluntary rhythmic leg movement frequency
Sardroodian M, Madeleine P, Voigt M, Hansen, EA
Board 14
Are intraoperative forces higher than the postoperative forces in scoliosis
corrective surgery? Remel Alingalan Salmingo, Shigeru Tadano, Yuichiro Abe,
Manabu Ito
Board 15
Multi-View Stereo Approach to Gait Analysis
Martin S. Christiansen, Henrik Aanæs, Thomas B. Moeslund, Henrik Koblauch, Tine
Alkjær, Erik B. Simonsen
Board 16
Temporal structure of lower limb kinematics during walking at different
velocities Peter C. Raffalt, Martin Kjær Guul, Andreas Nexmann Nielsen, Erik B.
Simonsen and Tine Alkjær
Board 17
The influence of gait speed on local dynamic stability and variability during
walking in transfemoral amputees
Magdalena Rós Guðnadóttir, Erik B. Simonsen, Tine Alkjær
U NI V E R S I T Y
O F
CO P E NH A G E N
Participants
Name
Tine Alkjær
Marius Henriksen
Jens Aaboe
Andreas Top Adler
Kevin Backes
Jesper Bencke
Daniel Benoit
Hanne Bloch
Finn Bojsen-Møller
Jens Bojsen-Møller
Martin S Christiansen
Helle Dalsgaard
Michel Dalstra
Mark de Zee
Rune Eie
Teresa Flaxman
Alba Granados
Magdalena Rós Guðnadóttir
Ernst Albin Hansen
Thomas Heilskov-Hansen
Frederik Heinen
Anders Holsgaard
Troels Kardel
Uwe Kersting
Henrik Koblauch
René Korsgaard
Mathias Kristiansen
Pascal Madeleine
Bjørn Gliese Madsen
Lars Michler
Pia Bondorph Nielsen
Peter C. Raffalt
John Rasmussen
Remel Alingalan Salmingo
Afshin Samani
Mahta Sardroodian
Sotiris Sarros
Erik B. Simonsen
Joakim Bay Simonsen
Tina Skytte
William Sloth
Karen Søgaard
Rasmus Skov Sørensen
Email
[email protected]
[email protected]
[email protected]
[email protected]
[email protected]
[email protected]
[email protected]
[email protected]
[email protected]
[email protected]
[email protected]
[email protected]
[email protected]
[email protected]
[email protected]
[email protected]
[email protected]
[email protected]
[email protected]
[email protected]
[email protected]
[email protected]
[email protected]
[email protected]
[email protected]
[email protected]
[email protected]
[email protected]
[email protected]
[email protected]
[email protected]
[email protected]
[email protected]
[email protected]
[email protected]
[email protected]
[email protected]
[email protected]
[email protected]
[email protected]
[email protected]
[email protected]
[email protected]
U NI V E R S I T Y
Ilias Theodorakos
Maj Halling Thomsen
Steffen Vangsgaard
Christian Wong
Mette K.Zebis
O F
CO P E NH A G E N
[email protected]
[email protected]
[email protected]
[email protected]
[email protected]
U NI V E R S I T Y
O F
CO P E NH A G E N
ABSTRACTS
Motor control during walking in
patients with knee osteoarthritis
Helle Dalsgaard, Tine Alkjær, Erik B. Simonsen, Marius
Henriksen
Purpose: The main purpose of this study was to quantify the motor control
in a group of people with knee osteoarthritis and an asymptomatic group. It
was investigated if there was a significant difference in Hoffmann(H-) reflex modulation compaired to the underlying neuromuscular activity between the two
groups. Movement variability was quantified too.
Method: Eleven patients with knee osteoarthritis and eleven asymptomatic
control subjects walked on a treadmill at 3.5 km/h. The experiment was divided into two parts: 1) Soleus H-reflex modulation, 2) EMG, as well as gait
parameters. H-reflex was recorded to reveal the underlying motor strategy in
walking in each of the groups. Data collection of electromyography (EMG),
angles of knee and ankle and signales from electrical foot switches under heel
and toe, took place during 6 minutes of walking. EMG for soleus (SO), tibialis anterior (TA), vastus medialis (VM) and biceps femoris (BF) muscles was
recorded. It was applied to calculate co-activation between SO and TA, as well
as VM and BF. Based on the recorded joint angles and data from foot switches,
the movement variability was quantified. This calculation of the coefficient of
variation was determined of stride time intervals which was found from a switch
under the heel, as well as the approximate entropy (ApEn) in joint angels were
calculated.
Results: The examination of the H-reflex using mixed model showed a significant difference between the groups in the first two intervals after heel strike,
and an increased H-reflex by OA Group at the end of the swing phase. This difference could not be explained by soleus EMG’et. On the contrary some of the
difference could be explained on the basis of TA-EMG that showed significance
in the intervals just about heel strike (Interval 1 (p = 0.0145) and interval 16 (p
= 0.0016) for the normalized EMG). There was no significant difference between
the groups in the moverment varibility either by calculating the coefficient of
variance of step time intervals or when calculating approximated entropy of the
joint angles. There was not found any difference in the other investigated time
parameters.
Conclusion: It was concluded that there were only differences between the
groups in H-reflex modulation around heel strike. Similar differences in the Hreflex pattern was previously found in two healthy groups of individuals. This
might indicate that the OA patients were using a motor control, which relieved the knee. There was not found any differences between the goups in the
movement variabilty and the remaining parameters studied.
1
A comparison of accuracy and precision of 5 gait-event detection algorithms
from motion capture in horses during over ground trot
Jenny Katrine Boyea,*, Emil Olsena,b, Thilo Pfaub, Maj Halling Thomsena
a: University of Copenhagen, Faculty of Health and Medical Sciences, Department of Large
Animal Sciences, Hoejbakkegaard Allé 5, 2630 Taastrup, Denmark.
b: The Royal Veterinary College, Department of Veterinary Clinical Sciences, Hawkshead Lane,
North Mymms, AL9 7TA, United Kingdom.
* Corresponding author. +45 2010 4229
E-mail address: [email protected] (J. K. Boye)
Abstract
Motion capture is frequently used over ground in equine locomotion science to
determine gait events (hoof-on/off and stance) without force plates. Robust and
validated algorithms for trot as well as comparative evidence for available methods
are lacking. Recent developments of motion capture towards outdoors gait analysis
without force plates emphasise the need for developing accurate algorithms. It is the
objective of this study to compare accuracy (bias) and precision (SD) for three human
and two equine published gait event detection algorithms for motion capture. Force
plates are used as reference standard. Six horses were trotted over 8 seamlessly
embedded force plates (Kistler Instrumente, AG, Switzerland) surrounded by a
synchronised 12-camera infrared motion capture system (Qualisys AB, Sweden). The
force plates were sampling at 1000 Hz and the motion capture at 240 Hz. Reflective
markers were placed over the right and left centre of mass, lateral hoof, proximal
dorsal hoof, distal dorsal hoof, fetlock joint and laterally on the proximal metacarpus
II / metatarsus II on all four legs. Algorithm A is based on a threshold in horizontal
hoof velocity [1]. Algorithm B was proposed as an improvement of A where hoof-off
is obtained using a threshold for the angle of the fetlock joint during flexion [2].
Algorithm C and D are based on hoof position and velocity relative to the Centre of
Mass position. For Algorithm C, hoof-on is defined by the maximum peak and hoofoff by the minimum peak in relative horizontal position of the hoof. In D the zero
crossing of relative horizontal hoof velocity defines hoof-on/off [3]. Algorithm E is
based on the zero crossing of vertical hoof acceleration for hoof-on and hoof-off is
defined by the first positive peak after stance [4].
Post processing of data was conducted using automated custom written scripts
in MATLAB (Mathworks, USA). Statistics were carried out using R (R Core
Development Team, r-project.org, version 2.13.1) applying agreement methods
according to Bland and Altman [5] corrected for repeated measures on each horse.
The data were analysed using the package MethComp with replicate measures for
front limbs or hind limbs. The best single performing algorithm was A in the front
limbs. Method A had an accuracy of 16.5 ms for hoof-on and -0.9 ms for hoof-off.
The precision (standard deviation) was 14.5 ms for hoof-on and 25.1 ms for hoof-off.
Method E was more precise than method A on hoof-off with an accuracy of 2.0 ms
and precision of 8.1 ms.
Combining A (hoof-on) and E (hoof-off) in the front limbs and B (hoof on and
off) in the hindlimbs lead to accuracy ranges between -2.0 and 19.2 ms and the
precision between 8.1 and 27.6 ms.
In conclusion, we provide the first comparison of all five methods for use in
over-ground trot in horses and provide evidence for a combination of published
algorithms to accurately and precisely detect gait events with motion capture.
Keywords: Equine, Locomotion, Method comparison, Gait events, Stance
References:
1.
2.
3.
4.
5.
Peham, C.; Scheidl, M.; Licka, T., Limb locomotion - speed distribution analysis as a new
method for stance phase detection. J Biomech 1999, 32, (10), 1119-24.
Galisteo, A. M.; Garrido-Castro, J. L.; Miró, F.; Plaza, C.; Medina-Carnicer, R., Assessment
of a method to determine the stride phases in trotting horses from video sequences under field
conditions. Wien. Tierärztl. Mschr. 2010, 67, (3 & 4), 65 - 73.
Zeni, J. A., Jr.; Richards, J. G.; Higginson, J. S., Two simple methods for determining gait
events during treadmill and overground walking using kinematic data. Gait Posture
2008, 27, (4), 710-4.
Leitch, J.; Stebbins, J.; Paolini, G.; Zavatsky, A. B., Identifying gait events without a force
plate during running: A comparison of methods. Gait Posture 2010.
Bland, J. M.; Altman, D. G., Agreement between methods of measurement with multiple
observations per individual. J. Biopharm. Stat. 2007, 17, (4), 571-82.
WEIGHT BEARING KNEE JOINT STABILISATION STRATEGIES DIFFER BETWEEN YOUNGER,
OLDER, AND OSTEOARTHRITIC POPULATIONS
1
2
1,2
Teresa E. Flaxman, Andrew J.J. Smith and Daniel L. Benoit
School of Rehabilitation Sciences, University of Ottawa, Canada
2
School of Human Kinetics, University of Ottawa, Canada; contact: [email protected], website:
dlbenoit.ca
1
Introduction
Osteoarthritis (OA) is the most debilitating condition among older adults. OA is mechanically driven
by altering the stabilising integrity of the joint [1]. The main contributor to knee joint stability is
muscular contractions. In cases where a traumatic knee joint injury is not a causal factor [2], a
change in muscle function, resulting in reduced strength and force control in believed to induce OA
development and progression. Since age is also a determining factor of OA, the purpose of this
study was to investigate the muscle activation patterns of young healthy adults (YC), older healthy
adults (OC), and adults with OA during a standing isometric force control task.
Patients & Methods
A force matching protocol [3] was used to evaluate muscle activation patterns of 41 YC (23.1±1.9
years of age) 18 OC (59.7±5.14 years), and 19 OA (63.5±8.1 years). Subjects stood with their leg of
interest fixed to a force platform and modulated ground reaction forces while exposing equal body
weight to each leg. Surface electromyography (EMG) of 8 muscles that cross the knee joint, kinetics
and kinematics were recorded while subjects generated 30% of their maximal force in 12 different
directions, corresponding to various combinations of medial-lateral-anterior-posterior ground reaction
forces.
Processed EMG was normalized to previously recorded maximum voluntary isometric contraction
(MVIC) and ensemble averaged into group means for each loading direction. Muscle activation
patterns were displayed in EMG polar plots and were quantified with symmetry analyses, mean
activation levels (XEMG), directions (Φ), and specificity indices (SI) [3]. Group differences were tested
with independent T-tests at the p<0.05 level.
Results
Muscle activation patterns were similar between all groups (i.e. symmetry and Φ) (Fig. 1 and 2).
However, XEMG was significantly greater in both the OA and OC groups compared to YC (Fig. 2A).
OA group also demonstrated significantly greater X EMG in the rectus femoris and tensor fascia lata as
well as lower SI in semitendinosus hamstrings (Fig. 2B) compared to OC.
Discussion/Conclusion
Our results indicate that regardless of loading direction, both OC and OA groups have greater levels
of muscle co-contraction than YC. This is suggested to be an adaptive response to age-related
changes in muscle strength and force control [4]. Since individuals with OA have reduced muscle
strength and force control compared to age-matched controls [4], our results suggest that the OA
group’s greater, less specific activation of knee joint muscles relative to the OC is this “stiffening”
response adapted by the OA group, however, to an extent that may expose the joint to detrimental
loading conditions, contributing to the progression of OA [5]. Further investigation regarding agerelated neuromuscular changes and joint loading conditions on development of OA is warranted.
References
1. Herzog W. (2006) Biomechan Model Mechanobiol. 5: 64–81.
2. Slemenda C. (1997) Ann of Intern Med. 127: 97-104.
3. Flaxman TE. (2012) J Biomech. 45:2570-6.
4. Rudolph KS. (2007) Phys Ther. 87:1422-32.
5. Herzog W. (2007) J Biomech. 40:S54-63.
Figure 1: Mean EMG polar plots of YC, OC, and OA groups. Numbers along the circular trajectory
represent the force direction in degrees. EMG magnitudes indicated by numbers along the vertical
radius where 1.0 is 100% MVIC and 0 is 0% MVIC.
Figure 2: (A) Mean direction of muscle activation (Φ) in degrees (°), (B) Mean magnitudes of muscle
activation (XEMG) and (C) Mean specificity indices (SI) of YC, OC, and OA groups. Values are
normalized from 0 to 1. Error bars indicate standard deviation. Letters (a,b,c) indicate significant
between group difference (p<0.05). a –OA and YC; b-OA and OC; c-OC and YC.
FE model of the hip based on clinically recorded CT
Background
In previous studies finite element (FE) models of the hip joint have been
constructed and validated. However, all of those models were created either from
post mortem collected data or with a contrast agent injected in the hip capsule
and has therefore little relevance in the clinic.
While most other studies have been concerned with the hip of adults or even
elderly patients, some studies suggest, that abnormalities in the hip joint during
childhood significantly increases the risk of developing osteoarthritis later in life. It is therefore relevant to investigate the hip joint already in childhood. Purpose/aim of the study
The aim of this study was to construct a FE model of a childʼs hip joint from an
ordinary clinical CT scan. Because previous studies have been concerned with
the pelvis of adults a direct comparison is not possible, but the validity of the new
model is discussed with regard to previous findings. Hopefully this model will
prove a useful tool for preoperative evaluation of surgical procedures in children.
Materials and methods
Due to a subluxated hip, a CT scan of the pelvic region on a 10-year old girl was
recorded. The healthy side was deemed normal and was used as the basis for
the FE model. In all previous studies the cartilage layers were segmented from
the CT, but in this clinically recorded CT the contrast between the cartilage and
the surrounding tissues were to poor. Therefore only the bones were segmented
from the CT. Cartilage was instead modeled as a hyper-elastic interaction
between the bone of the acetabulum and the femoral head and was assigned a
constant thickness over the entire surface. The model was assigned boundary
and loading conditions to imitate previous studies that combined FE models and
experimental analysis.
Findings/results
Firstly the distribution of von Mises stress in the cortical bone of the pelvis was
considered. While others have reported stresses of up to 6 MPa, we found
regions of higher stresses reaching 10 MPa. Regions of highest stresses are in
accordance with one of the previous studies, while another study revealed a
different distribution of the stresses.
Secondly, the model reported cartilage contact pressure in the acetabulum of
around 0-10 MPa. This is in accordance with previous findings. Regions of peak
forces are slightly different, but within what are considered acceptable, since
large variations even among healthy subjects have been reported previously.
Conclusions
We have developed a FE model of the hip joint from an ordinary preoperatively
recorded CT scan. Compared to previous models, this simpler model reveals
contact forces in the acetabulum that is deemed adequate. Further works are
however still before the model can be used in clinical applications. DOES QUADRICEPS-STRENGTHENING EXERCISE AFFECT QUADRICEPS FORCE, POWER,
AND WORK DURING STAIR ASCENT IN ADULTS WITH KNEE OSTEOARTHRITIS?
Josh Leonardis1, Patrick Rider1, John Norbury2, Jens Aaboe3, Robin Christensen3,
Henning Bliddal3, Marius Henriksen3 & Paul DeVita1
1
2
Biomechanics Laboratory, Department of Kinesiology, East Carolina University, Greenville, NC
Department of Physical Medicine and Rehabilitation, BSOM, East Carolina University, Greenville, NC
3
Parker Institute, Copenhagen University Hospital, Frederiksberg, Denmark
email: [email protected], web: http://www.ecu.edu/cs-hhp/exss/biomechlab.cfm
INTRODUCTION
Osteoarthritis (OA) is a chronic joint disease that
accounts for more than half of all arthritis cases in
the US by affecting ~27 million people (4). OA is
characterized by joint space narrowing, exposure of
subchondral bone, irregular bone growth, and joint
pain and is most common in the knee (3, 4, 6). In
addition to joint pain and degradation, knee OA
causes the significant secondary problem of
quadriceps weakness. The effects of knee joint pain
and quadriceps muscle weakness are most evident
during activities of daily living, such as ascending
stairs, and can negatively affect quality of life.
Most treatment options for knee OA are limited to
targeting the symptoms of the disease and a cure
remains elusive. Many exercise interventions have
reduced pain while quadriceps-strengthening
exercise improves one’s physical function in
addition to the disease’s symptoms (1, 5, 7). The
mechanism behind improved symptoms, however,
has yet to be identified. It is widely believed that by
increasing quadriceps strength one improves
quadriceps function during locomotion, thus
reducing knee joint loads, pain, & disability.
However, no empirical evidence has been found to
support this theory. In fact, it has been found that
with an exercise-induced increase in quadriceps
strength, negligible changes occur to knee joint
mechanics during locomotion, even with a cessation
in pain (1-2). Therefore, we hypothesize that
quadriceps-strengthening exercise will not affect
quadriceps muscle biomechanics during locomotion
in adults with knee osteoarthritis. The purpose of
this pilot study was to determine the effects of a 12week quadriceps strengthening protocol on
maximum quadriceps muscle force, power, and
work during stair ascent in adults suffering from
knee OA.
METHODS
10 adults with physician-diagnosed tibiofemoral
knee OA volunteered and were randomly placed
into a trained or untrained group after providing
written informed consent. Their characteristics
were:
Table 1: Age, height, weight, and BMI.
Trained
Untrained
6
4
Number
54.3 ± 5.8
57.4 ± 4.7
Age (yrs)
1.7 ± .04
1.7 ± .10
Height (m)
77.8
±
8.9
76.0
± 21.5
Mass (kg)
2
27.2 ± 3.2
26.9 ± 4.4
BMI (kg/m )
Participants had kinematic, ground reaction force,
isokinetic strength, and WOMAC data collected
before and after a 12-week interval. During this
interval the trained group participated in a
quadriceps-strengthening protocol 3 sessions per
week, while the untrained group were not enrolled
in any activity program. Training sessions consisted
of 3 exercises performed for 3 sets of 10 repetitions
with loads increasing from 65% to 80% of a
previously determined 3RM. Quadriceps muscle
force, power, and work were quantified using
kinematic and kinetic data in combination with a
biomechanical knee model (5, Fig. 2).
T-tests were used to determine significant pre- and
post-test differences in strength and WOMAC data,
while a 2-way repeated measures ANOVA was used
to determine differences in quadriceps muscle force,
power, and work between pre- and post-tests and
treatment groups, all using p<0.05.
that the pain relief mechanism cannot be explained
solely by improved quadriceps biomechanics.
Figure 2. Biomechanical model
Figure 3. Quadriceps force and power in stair
ascent; solid & dotted: pre-test mean±sd; dashed:
post-test mean
RESULTS AND DISCUSSION
Consistent with the literature, the training group
exhibited significant increases in quadriceps
strength as well as improvements in pain, function,
and total WOMAC scores (all p<0.05). The control
group showed no changes in strength or symptoms.
Maximum quadriceps force and power (Fig. 3) were
unchanged during ascent in both groups. These data
are highlighted in Table 2 for the strength-training
group.
CONCLUSIONS
Despite improvements in strength, pain, and
function, quadriceps-strengthening exercise did not
change quadriceps force, power, or work during
stair ascent. The findings of this pilot study suggest
REFERENCES
1.Foroughi N, et al. Cl Biomech 26, 167-174, 2011
2.Hurwitz DE, et al. J Orth Res 18, 572-579, 2005
3.Lawrence RC, at al. Arth & Rheum 58 (1), 26-35,
2008
4.London NJ, et al. Med Hyp 76, 887-892, 2011
5.Messier SP, et al. Ost & Cart 19, 272-280, 2011
6.Vos T, et al. Lancet 380, 2163-2196, 2010
7.Wang TJ, et al. Cl Nursing 20, 2609-2622, 2010
Table 2: Strength, symptoms, and quadriceps muscle biomechanics in strength training group.
Pain
Function
Total
Strength
Force
Power
Work
3.5 ± 4.2
10.0 ± 12.2
20.9 ± 23.9 102 ± 41 Nm
28 ± 4.3 N
2.5 ± 0.5 W 0.7 ± 0.1 J
Pre
1.0 ± 1.3*
8.6 ± 13.4*
Post 1.5 ± 2.2*
*Significant pre- post-test differences (p<.05)
133 ± 32 Nm*
31 ± 4.8 N
2.7 ± 0.4 W
0.8 ± 0.1 J
poster sessions
APPLICATION OF ERGOVAR, AN INTERACTIVE OPENSOURCE TOOLBOX IN MATLAB, IN BIOMECHANICS
Afshin SAMANI
Center for Sensory-Motor Interaction (SMI), Dept. of Health Science and Technology, Aalborg
University, Aalborg, Denmark, [email protected]
Corresponding author e-mail: [email protected]
Abstract
Purposes
It has been hypothesized that increased exposure variation may prevent work-related musculoskeletal
disorders development in jobs characterized by long-term exposure to constrained postures or
repetitive movements. Such assumption calls for the development of methods that can quantify
biomechanical exposure variation within and between subjects at work. Despite the existence of
computational methods and potential relevance, recurrent patterns throughout a time-line of exposure
have only been rarely investigated.
Thus, we aimed at developing an open-source toolbox in MATLAB that allows quantification of
exposure variation. The goal was to develop this toolbox named ErgoVar using the MATLAB main
core so that researchers could easily get access to the codes and apply them to their own application.
Methods
ErgoVar provides a collective framework offering processing tools to the researcher in the field of
ergonomics. ErgoVar is based on a user-friendly graphical interface which presents the processing
tools to the users who are not familiar with MATLAB. The current version of ErgoVar takes a minimal
benefit of a few common MATLAB toolboxes such as signal processing toolbox. However, the ultimate
aim is to promote a stand-alone application in MATLAB environment. The removal of ErgoVar
dependencies to the MATLAB toolboxes is feasible but those toolboxes speed up the running time.
ErgoVar is now available online at (http://smi.hst.aau.dk/research/projects/ergovar/). ErgoVar is
provided "AS IS" and expressly disclaims all warranties. It can only be share and modified for
research.
Results
The ErgoVar can import user’s data or simulate synthetic data and work with them. At the current
version, ErgoVar provides tools to compute entropy measures (sample, permutation and multi-scale
entropies), recurrence quantification analysis of nonlinear dynamics, exposure variation analysis and
a newly developed method called cluster-based exposure variation analysis.
Conclusions
ErgoVar is still at its early stage of development and the plan is to make it as broad as possible
including different methodologies of relevance to exposure variation analysis. ErgoVar is distributed
under the GNU General Public License (version 3 or later). ErgoVar is financially supported by the
Danish Council for Independent Research | Technology and Production Sciences (FTP). Grant
number: 10092821.
Keyword : Nonlinear processing, Repetitive tasks
October 25th | Copenhagen
Non-negative matrix factorization algorithm as a way to analyze muscle
synergies: Application to bench press exercise
Mathias Kristiansen; Pascal Madeleine; Ernst Albin Hansen; Afshin Samani
Center for Sensory-Motor Interaction (SMI), Department of Health Science and Technology,
Aalborg University, Denmark
The purpose of the study was to develop and apply a decomposition algorithm based on nonnegative matrix factorization to assess muscle synergies during bench press. Ten male expert power
lifters (EXP) completed 3 sets of 8 repetitions at 60% of 3 repetition maximum in bench press.
Muscle synergies were extracted from surface electromyography (EMG) data from 21 bench press
cycles using the non-negative matrix factorization algorithm (1). After finding the minimum
number of muscle synergy vectors and synergy activation coefficients which described more than
90% of the total variability, the muscle synergies extracted from each subject were functionally
sorted (2). The functional sorting was done to ensure that the order of the extracted muscle
synergies was maintained for all subjects. The muscle synergies were ordered based on the
similarity of their muscle synergy vector and/or synergy activation coefficients to those of an
arbitrary reference subject. The muscle synergy vector reflects the contribution of single muscles
within a muscle synergy while the synergy activation coefficient reflects the strength of a muscle
synergy within the entire movement. Two muscle synergies described more than 90% of the total
variability and reflected the eccentric and concentric phase. Before the functional sorting was
applied, muscle synergy 1 represented the concentric phase of the bench press, while muscle
synergy 2 represented the eccentric phase. After applying the functional sorting, muscle synergy 1
represented the eccentric phase of the bench press, while muscle synergy 2 represented the
concentric phase. Further, cross correlations for synergy activation coefficient 1 and 2 were 0.65
[0.52;0.79] and 0.59 [0.49;0.77] (Median rmax [25th;75th percentile]). The present results indicate
that a functional sorting of the muscle synergies is central for the interpretation.
References
(1) Hug F, Turpin NA, Couturier A, Dorel S. Consistency of muscle synergies during pedaling
across different mechanical constraints. Journal of Neurophysiology 2011 July 01;106(1):91-103.
(2) Torres-Oviedo G, Ting LH. Muscle synergies characterizing human postural responses. J
Neurophysiol 2007;98(4):2144-2156.
5th annual DBS meeting
October 25 2013, Copenhagen, DK
th
Ankle bracing changes lower extremity alignment during landing
Ilias Theodorakos 1, Jan Rueterbories 1, Michael S. Andersen 2, Mark de Zee 1 and Uwe G. Kersting 1
1
Department of Health Science and Technology, Aalborg University, Denmark
Department of Mechanical and Manufacturing Engineering, Aalborg University, Denmark;
E-mail: [email protected], web: www.smi.hst.aau.dk
2
ABSTRACT
An inversion trauma causes 85% of ankle sprains, and ankle ligament injuries constitute about 25% of the injuries that
occur in running and jumping sports [1]. Baumhauer and colleagues have reported that the combination of a more
inverted and plantarflexed positioning of the foot at touchdown increases the risk of a lateral ankle sprain [2]. In order to
reduce the incidence of ankle sprains, ankle bracing is commonly used. Trap doors, cutting maneuvers, and landing
protocols have been used to test ankle braces under dynamic conditions. Recently, Chen and colleagues reported that
landing on inclined surfaces is more demanding than trap door mechanisms and thus better for investigating ankle braces
and lateral ankle performance mechanisms [3].
Sixteen healthy men with no prior ankle injuries participated in this study. Participants were asked to perform a landing
task from a height of 40 cm on a robotic force platform [4]. The robotic platform was randomly inclined at four different
angles during the flight phase, producing 5° eversion and 0°, 10° and 15° inversion when making contact with the plate.
Participants were instructed to perform the task by using their dominant leg and starting from a defined position. Landings
were repeated seven times per inclination with and without an ankle brace. A trial was considered successful when the
participants were able to land with their testing leg on the force plate and maintained their balance on that leg for at least
2 s. Kinematic variables obtained by a five segments stick-figure model of the lower extremity [5], ground reaction forces,
and electromyographic activity of five lower extremity muscles were analyzed. Two-way (4 inclined surfaces x 2 brace
conditions) repeated measures ANOVAs tested for group differences with significance value set to p<0.05 for all analyses.
No significant differences were observed for ground reaction forces. Tibialis anterior (TA) muscle activity was significantly
increased while the muscle activity of lateral gastrocnemius (GL) was significantly decreased before touchdown for the
braced condition. Significant differences were observed for the ankle and knee joints at touchdown. The talocrural joint
was less plantarflexed, the subtalar was more everted, and the knee was more flexed for the braced condition. It seems
that bracing changed the pre-activation pattern. Tricept surae muscle was less activated while TA showed increased
activity. This different pattern resulted in different landing positions/strategies. A more everted and less plantarflexed
positioning of foot at touchdown decreases the risk of a lateral ankle sprain [2] while a more flexed knee functions as a
better shock absorber [6]. The results showed that bracing provides a more safe and stable positioning of the leg at
touchdown and thus bracing decreases the risk of an ankle sprain during landing.
REFERENCES
1.
2.
3.
4.
5.
6.
Bahr, R., et al., Incidence and mechanisms of acute ankle inversion injuries in volleyball. A retrospective cohort study. Am J
Sports Med, 1994. 22(5): p. 595-600.
Baumhauer, J.F., et al., A prospective study of ankle injury risk factors. Am J Sports Med, 1995. 23(5): p. 564-70.
Chen, Q., et al., Is the inverted surface landing more suitable in evaluating ankle braces and ankle inversion perturbation? Clin
J Sport Med, 2012. 22(3): p. 214-20.
van Doornik, J. and T. Sinkjaer, Robotic platform for human gait analysis. IEEE Trans Biomed Eng, 2007. 54(9): p. 1696-702.
Lund, M.E., et al., Functional Scaling of Musculoskeletal Models, in Congress of the International Society of Biomechanics, ISB.
2011: Brussels, Belgium.
Nagai, T., et al., Knee proprioception and strength and landing kinematics during a single-leg stop-jump task. J Athl Train, 2013.
48(1): p. 31-8.
Using isokinetic dynamometer measurements to determine muscle strength
characteristics of elite cyclists: a preliminary study
1
Frederik Heinen, 2Carsten Møller Mølgaard, 3John Rasmussen and 1Mark de Zee
1
Department of Health Science and Technology, Aalborg University, Denmark
2
University College Nordjylland, Denmark
3
Department of Mechanical and Manufacturing Engineering, Aalborg University, Denmark
email: [email protected]
INTRODUCTION
The Hill-type muscle model has been shown to be very
Isokinetic measurements were conducted for the same joint
sensitive towards the tendon slack length (Lt0) and optimal
movement as in the isometric measurement at different
muscle fiber length (Lf0) [1]. These muscle parameters are
angular velocities (table 2). The isokinetic measurements are
often based on data from morphologic studies, however this
performed both eccentric and concentric at the given
pose some issues when trying to analyze muscle forces from
velocities for both joint movements. The angular velocities
subjects that deviate from the specimen. Van den Bogert et
were chosen based on the reported joint angle velocities that
al. 1998 suggested that a “hybrid” approach should be
occur during the cycling movement [5]. Three consecutive
implemented in such a manner that some muscle properties
repetitions were performed at each angular velocity.
are derived from morphological data while others by fitting
the model to subject-specific data [2].
Placement and fixation of the subject for the different
It has been shown that the strength of musculoskeletal
measurements followed the recommendations of the
models based on cadaveric data can deviate from
manufacture. Axis alignment between the joint axis and
experimental data and needs to be tweaked in order to
dynamometer was performed based on bony landmarks.
produce reliable results [3].
A non-invasive method for obtaining muscle force
DISCUSSION AND EXPECTED OUTCOME
parameters in vivo is by means of an isokinetic
It has been shown that care should be taken aligning the
dynamometer which measures the torque produced by the
joint axes and the dynamometer since misalignment will
muscles operating a certain joint movement. This method
affect the joint torque production of the subject [6]. It can
has widely been used to obtain muscle force parameters
therefore be argued weather the experimental protocol
such as peak isometric force, force-velocity relation and
should be executed in such a manner, that the subject is not
force-length relations of different muscle groups [4].
moved during the measurement but rather that one joint at
one leg should be finalized before changing sides. However,
The aim of this study is therefore to develop a protocol to
this was not possible for practical reasons, since this would
collect muscle strength data from elite cyclists using an
require a lot longer experimental duration due to the breaks
isokinetic dynamometer that can be implemented by means
between the isokinetic and isometric tests. Therefore, special
of mathematical optimization in a musculoskeletal model.
care was taken to make sure that the axes alignment was
performed the same way. Furthermore, test re-test studies
have shown that there is no significant difference between
EXPERIMENTAL PROTOCOL
Elite male mountain bike cyclists participated in this study.
different measurements if an appropriate protocol is used
The subjects’ performed a series of isokinetic and isometric
[7].
measurements in an isokinetic dynamometer (Biodex 4,
Biodex Medical Systems).
The expected outcome of this experiment is that the
collected dynamometer data can be used to create subjectIsometric measurements were conducted for the ankle
specific musculoskeletal models and guidelines for how this
plantar/dorsi flexor muscles, knee and hip flexor/extensor
should be performed.
muscles at different angles (table 1). The isometric
contraction was performed for four seconds and repeated
REFERENCES
three times for each joint angle.
1. De Groote, et al. Journal of Biomechanics (10): 18761883, 2010
2. van den Bogert, et al. Journal of Electromyography and
Table 1: Isometric joint angles. The same angels are used for both joint
Kinesiology: 119-124, 1998
movements.
3. Lund, et al. International Society of Biomechanics: 11Joint movement
Joint angle (°)
12, 2013
Ankle plantar/dorsi flexion
15, 5, 0, -5, -15
4. King, & Yeadon, M. R. Journal of applied
Knee flexion/extension
30, 60, 90, 110
biomechanics: 207-217, 2002
Hip flexion/extension
30, 60, 90. 110
5. HULL, & JORGE, M. Journal of Biomechanics: 631644, 1985
Table 2: Isokinetic angular velocities. The same angular velocities are
6. Rothstein, et al. Journal of Physical Therapy: 1840used for both concentric and eccentric muscle contraction.
1844, 1987
7. McCleary, & Andersen, J. C. Journal of Athletic
Joint movement
Angular velocity (°/s)
Training: 362-365, 1992
Ankle plantar/dorsi flexion 60, 120, 180. 210
Knee flexion/extension
60, 120, 180. 240
Hip flexion/extension
60, 120, 180, 240
LUMBAR LOAD IN COMMON WORK TASKS FOR AIRPORT BAGGAGE HANDLERS – A 3D DYNAMIC
ANALYSIS
1
Koblauch H, 2Bern S.H., 2Brauer C, 2Mikkelsen S, 3de Zee M, 4Thygesen LC, 4Helweg-Larsen K, 1Alkjær T, 5Andersen MS,
1
Simonsen EB
1 Department of Neuroscience and Pharmacology, University of Copenhagen, Copenhagen, Denmark;
2 Department of Occupational and Environmental Medicine, Bispebjerg University Hospital, Copenhagen, Denmark;
3 Department of Health Science and Technology, Aalborg University, Aalborg, Denmark;
4 National Institute of Public Health, University of Southern Denmark, Copenhagen, Denmark;
5 Department of Mechanical and Manufacturing Engineering, Aalborg University, Aalborg, Denmark;
Contact: Henrik Koblauch. [email protected]
SUMMARY
Baggage handlers often have a history of work-related low
back pain. The baggage handlers often work in confined
spaces, necessitating sitting, kneeling or stooping positions.
These positions have been associated with low back pain.
The purpose of this study was to quantify the lumbar
loading in airport baggage handlers with the help of
musculoskeletal modeling.
We selected two common work tasks (standing baggage
handling and sitting position inside the luggage
compartment). A single subject performed the tasks, while a
custom build motion capture system with eight synchronized
high-speed cameras recorded 3D kinematics. Two force
platforms recorded the ground reaction forces under the feet
and one under pelvis during the sitting position. We used
The AnyBody Modeling System v. 5.3 to build a model and
to compute the compression forces of the lumbar spine
(L4/L5-segment).
The standing position produced a peak L4/L5-compression
of 3980 N and the sitting position produced 2600 N.
This study described the loading of the lumbar spine during
common work tasks for airport baggage handlers. The level
of spinal peak compression in the standing task exceeded the
recommendation proposed by NIOSH. However, the
NIOSH recommendations are based on static measurements
and do not allow movement or asymmetrical lifting i.e. onehanded lifts or body rotations.
INTRODUCTION
Musculoskeletal pain constitutes a major occupational
health hazard in highly developed industrialized societies.
This is not only a problem for the individual worker; this is
also a massive economic issue. In 2011 the cost of back-pain
alone amounted to 1.7 billion Euros in Denmark. Around
35% of the Danish population has low back pain, and 13 %
of those with long-term illness have back pain.
Approximately 60 % of airport baggage handlers report that
they have had low back pain in the last year [1]. The job as
an airport baggage handler consists of frequent lifting. The
baggage handler lifts suitcases with a mean of 15 kg,
resulting in a mean daily accumulated load of around 5 tons
[2]. Further, baggage handlers often work in confined spaces
in the luggage compartments of aircrafts, necessitating
sitting, kneeling or stooping positions. Non-neutral working
positions and lifting are important factors in developing low
back pain [3]. Further, working in confined space has been
shown to increase the load on the lumbar spine [4].
Increased compression forces in the lumbar spine have also
been suggested as a risk factor for developing low back pain
[5]. The purpose of this study was to quantify the lumbar
loading in airport baggage handlers with the help of
musculoskeletal modeling.
METHODS
We selected two common work tasks for the airport baggage
handler: 1) Sitting position in the luggage compartment and
2) the task of loading baggage from the baggage cart to the
conveyor in a standing position. These constituted a main
part of work tasks for a baggage handler on the apron.
3D kinematics was obtained by a custom made motion
capture system of eight synchronized high-speed HD
cameras. A single subject (male, 30 years, 1.69 m., 65 kg.)
wearing a full body marker setup of 38 markers was
recorded while performing the two tasks with a 15 kg
suitcase. Two force platforms (AMTI, Watertown, MA,
USA) recorded the ground reaction forces under the feet and
one under pelvis during the sitting position.
Inverse dynamics-based musculoskeletal models of the tasks
were built in the AnyBody Modeling system v. 5.3
(AnyBody Technology, Aalborg, Denmark) based on the
GaitFullBody model in the AnyBody Managed Model
Repository (AMMR) v. 1.5.
The peak compression forces in the L4/L5-segment were
extracted from the model.
RESULTS AND DISCUSSION
L4/L5 Compression force (N)
4000
3000
2000
1000
0
Sitting
Standing
Task
Figure 1. The peak compression force on the L4/L5segment. Standing position: 3891 N (61.2 N/kg*BW),
Sitting: 2600 N (40 N/kg*BW).
This study estimated the loading on the lumbar spine in two
different common work tasks for airport baggage handlers.
The standing position (3981 N) produced larger peak L4/L5compression force than the sitting position (2600 N) (Figure
1).
In the standing position (Figure 2) the peak compression
forces exceeded the NIOSH recommendations of 3400 N
[6]. The NIOSH equations, however, do not take restricted
workspace, working position as seated or kneeling or onehanded lifting into account. Further, these recommendations
are based on static 2D calculations where rotations and
accelerations are neglected. The present analysis was a 3D
dynamical analysis and may therefore have estimated the
loading of the body more accurately. Thus, this 3D dynamic
analysis can contribute to the extension of knowledge into
ergonomic musculoskeletal loading which typically has
relied on more simple 2D static analyses.
The sitting position (Figure 2) is only one of many possible
postures in the luggage compartment. Kneeling, stooped and
squatting are also commonly used positions [7]. Thus, these
results only represent a small part of the daily loadings
baggage handlers are exposed to inside the aircraft luggage
compartment. However, the standing position is a very
common position when handling luggage. It is observed that
the standing position is used when moving luggage from the
belly-cart to the conveyor by the aircraft and vice versa.
This position is also used inside the baggage terminal where
luggage is transferred into and out of belly-carts. Working in
standing position could constitute up to 50 % of a regular
work day for an airport baggage handler. Further, the task of
handling objects in a standing position can probably be
transferred to many other professions, so the results from
this study may also be valid in other aspects.
.
Figure 2: The model performing the standing and sitting
tasks.
CONCLUSIONS
In conclusion, the peak compression in the L4/L5-segment
was 3981 N (61.2 N/kg*BW) for handling baggage in a
standing position, while the peak compression was 2600 N
(40 N/kg*BW) in the sitting position. However, the
selection of work tasks was not exhaustive, so further
investigation must be carried out to fully describe the
loading of airport baggage handlers.
REFERENCES
1. Stalhammar HR, et al. Postural, epidemiological and
biomechanical analysis of luggage handling in an
aircraft luggage compartment. Appl Ergon, 17(3):177183, 1986.
2. Bern SH, et al. Unpublished data from the present study
of baggage handlers.
3. Griffith LE, et al. Individual participant data metaanalysis of mechanical workplace risk factors and low
back pain. Am J Public Health, 102(2):309-318, 2012
4. Harkness EF, et al., Risk factors for new-onset low back
pain amongst cohorts of newly employed workers.
Rheumatology (Oxford), 42(8):959-968, 2003
5. Bakker EW, et al. Spinal mechanical load as a risk
factor for low back pain: a systematic review of
prospective cohort studies. Spine, 34(8):E281-E293,
2009
6. Waters TR, et al. Revised NIOSH equation for the
design and evaluation of manual lifting
tasks. Ergonomics, 36:749-776, 1993
7. Stålhammar HR, et al. Postural, epidemiological and
biomechanical analysis of luggage handling in an
aircraft luggage compartment. Appl Ergon. 17(3):17783, 1986
High intensity eccentric training in the neck-shoulder region
Steffen Vangsgaard • Ernst Albin Hansen • Pascal Madeleine
Center for Sensory-Motor Interaction (SMI), Department of Health Science and Technology,
Aalborg University, Aalborg, Denmark
Introduction
Strength training has been suggested as a treatment for musculoskeletal disorders. Eccentric training is
known to be effective in the treatment of patients with shoulder problems. In addition, eccentric training is
associated with greater neural adaptation compared to concentric training. However, eccentric exercises have
not been sufficiently evaluated in relation to the neck-shoulder region. Thus, the objective of this study was
to investigate the effects of a training program based on eccentric exercises of the shoulder and neck muscles
on maximum strength, rate of force development, and electromyographic activity.
Methods
Twenty nine healthy subjects participated in the study (19 women and 10 men). The participants were
randomly divided into an eccentric neck-shoulder training group (ECC) and a reference group (REF). The
ECC program consisted of 9 sessions of eccentric exercise performed at 60-80% MVC over 5 weeks.
Maximal voluntary contraction (MVC), rate of force development (RFD) as well as electromyographic
(EMG) activity (voluntary and elicited), were recorded before (PRE) and after 5 weeks of eccentric training
(POST). MVC was calculated as the average of three maximal isometric shoulder elevations of 5 s of
duration. RFD was calculated from contraction onset to 500 ms.
Results
The training compliance was extremely high (100 %). As expected, this supervised intervention resulted in a
significant increase of 26.4 ± 20.9% from PRE to POST test in MVC force among the ECC group (P<0.01).
In accordance with this, the RFD increased significantly (24.6 ± 39.3 %) in the ECC group (P=0.025). EMG
data is currently being analyzed.
Discussion
This is the first study to investigate and document neuromuscular changes following eccentric strength
training of the neck-shoulder region. The knowledge obtained will be able to contribute to the treatment and
prevention of neck-shoulder pain in the future.
Sex differences in muscular load among house painters
performing identical work tasks
Thomas Heilskov-Hansen, MSc1, Jacob Meyland, MSc1, Tine Alkjær, PhD2, Henrik Koblauch
Baldvinsson, MSc2, Sigurd Mikkelsen, DMSc1, Susanne Wulff Svendsen, PhD3, Jane Frølund
Thomsen, PhD1, Gert-Åke Hansson, PhD4,5, Erik Bruun Simonsen, PhD2.
1
Department of Occupational and Environmental Medicine, Bispebjerg University Hospital,
Copenhagen, Denmark.
2
Institute of Neuroscience and Pharmacology, University of Copenhagen, Denmark.
3
Danish Ramazzini Centre, Department of Occupational Medicine, Herning Regional Hospital,
Herning, Denmark.
4
Division of Occupational and Environmental Medicine, Lund University, SE-221 85 Lund,
Sweden
5
Occupational and Environmental Medicine, Labmedicin Skåne, SE-221 85 Lund, Sweden
Correspondence to: Thomas Heilskov-Hansen, Department of Occupational and Environmental Medicine,
Bispebjerg University Hospital, Bispebjerg Bakke 23, DK-2400 Copenhagen NW, Denmark. Email:
[email protected] Phone: +45 35 31 37 88
Objective Aim of the present study was to investigate differences in muscular load between male
and female house painters performing identical work tasks, for the purpose of determine if sex
related differences, in respect of muscular activity, may help explain differences in occurrence of
MSD.
Methods In a laboratory 16 male and 16 female house painters, each performed nine standardised
work tasks common to house painters. Unilateral EMG recordings were obtained from the
supraspinatus muscle by intramuscular electrodes and from the trapezius-, extensor- and flexor
carpi radialis muscles by surface electrodes. Statistics were done using a mixed model approach on
10th, 50th and 90th percentiles. Both relative muscular load in %EMGmax as well as exerted force in
Newton, were assessed.
Results Regarding relative muscular load, sex was found to be a significant effect in m.
supraspinatus as well as in the forearm muscles, where women were exposed to a higher relative
load than men. Men exerted greater force in the trapezius muscle at the 50th percentile and in a
single task also the 10th percentile. Interaction between tasks and sex were in all but one single case
found to be non significant, indicating differences between sexes being independent of tasks.
Conclusion Female house painters are found to be subjective to a higher relative muscular load
than their male colleagues, even though men exert more force compared to women. The effects of
higher relative muscular load, accumulated over years of work, may in part explain why MSD in the
upper body occur more frequent among female house painters.
Key terms: Electromyography, emg, intramuscular emg, msd, muscoloskeletal disease, occupational health,
painter, surface emg.
HEAVY STRENGTH TRAINING AND VOLUNTARY RHYTHMIC LEG MOVEMENT FREQUENCY
Sardroodian M, Madeleine P, Voigt M, Hansen, EA
Department of Health Science and Technology, Aalborg University, Denmark
Introduction
Freely chosen cadence (FCC) during submaximal cycling has been suggested to represent an innate
voluntary rhythmic leg movement frequency, under primary influence of central pattern generators (Hansen
& Ohnstad, 2008). In other words, FCC may be a good reflection of central pattern generator movement
frequency output. Periods of heavy leg strength training, involving both hip flexion and extension exercise,
have been reported to cause recreationally active subjects to reduce their FCC during submaximal cycling
(Hansen et al., 2007; Rønnestad et al., 2012). The changed rhythmic movement behavior was observed as
early as after four weeks of training (Rønnestad et al., 2012). However, this change may even occur even
earlier, in response to neural adaptations. In this study, we investigated the early effect of separate
interventions with hip extension strength training and hip flexion strength training on voluntary rhythmic leg
movement frequency represented by FCC.
Methods
A total of 27 recreationally active individuals were randomized into three groups. Nine participants
constituted a group that performed heavy hip extension strength training (HET). Other nine participants
represented a group that performed heavy hip flexion strength training (HFT). In addition, nine participants
constituted a control group (CON) not exposed to any training. Training consisted of two sessions per week
for four weeks. In each session, three 5RM- to 10RM-sets were performed with each leg. Cycling was
performed before the start of the training period (baseline) and after every week of training as 5 min bouts at
100 W at FCC, on an SRM ergometer.
Results
Maximal strength (one repetition maximum, 1RM) increased as expected by 32±11% and 24±9% in HET
and HFT, respectively, after four weeks of training. These increases were larger than the change in CON
(p=0.001 and p=0.003, respectively). FCC was not different between the three groups at baseline (p=0.538).
Further, FCC in HFT and CON remained similar across the whole study period (p=0.633 and p=0.422). On
the other hand, percentage reductions of FCC in HET, as compared to baseline values, were larger in the
period from the second to the last test in the study period, as compared to corresponding values in CON,
(p=0.037). This was not the case when comparing CON and HFT (p=0.563).
Discussion
A novel finding of the present study was that a reduction of FCC appeared to occur early in a four-week
period of heavy hip extension strength training. Heavy hip flexion training did not have the same effect. We
propose that changes within the neuromuscular system caused the change in movement rhythm. A possible
hypothesis is that the hip extension strength training improved the efficacy of the neuromuscular system so
that the preset submaximal cycling could be performed with a reduced common descending supraspinal
drive. Such a modulation might cause a reduced stimulation of central pattern generators with a reduced
voluntary rhythmic leg movement frequency output as a consequence.
References
Hansen EA, Raastad T, Hallén J (2007) Eur J Appl Physiol, 101,419–426.
Hansen EA, Ohnstad AE (2008) Exp Brain Res, 186, 365-73.
Rønnestad BR, Hansen EA, Raastad T (2012) J Strength Cond Res, 26, 158-66.
Are intraoperative forces higher than the postoperative forces in scoliosis corrective surgery?
Remel Alingalan Salmingo a, Shigeru Tadano b, Yuichiro Abe c, Manabu Ito c
a
Biomedical Engineering, DTU Elektro, Ørsteds Plads, Building 349, 2800 Kgs. Lyngby
b
Laboratory of Biomechanical Design, Division of Human Mechanical Systems and Design, Hokkaido University,
N13 W8, Sapporo, 060-8628 Japan
c
Advanced Medicine for Spine and Spinal Cord Disorders, Hokkaido University, N15 W7, Sapporo, 060-8638 Japan
Patient 2
Fixation Level
Patient 1
Fixation Level
Abstract
Scoliosis is a complex pathology characterized as a three-dimensional (3D) deformity of the spine
combined with rotation of the vertebrae. Treatment for severe scoliosis is usually attained when the
scoliotic spine is deformed through application of corrective forces supported by the implant rods. It has
Convex Side
Concave Side
been previously established that the deformation of implant rod after scoliosis surgery is an inevitable
T5
T5
consequence of the postoperative corrective forces acting on the spine.[1,2]
However, the
intraoperative
T6
T6
deformation of implant rod has never been reported before. Thus, investigation on the changes
of implant
T7
T7
rod geometry (i.e. from intraoperative to postoperative) is necessary to understand the deformation
T8
behavior of implant rod during scoliosis surgery. This could help us T8understand the
magnitude of
T9
T9
intraoperative corrective forces acting on the spine and as well as the biomechanics of scoliosis
correction.
T10
T10
The main objective of this study was to calculate the intraoperative and postoperative corrective forces
T11
using Finite Element Analysis using the changes of implant rod geometryT11
during scoliosis
surgery.
T12
T12
Two rods were surgically implanted to the concave and convex side of the scoliotic spine (see below,
L1
L1
left and middle). Implant rod images were obtained intraoperatively and postoperatively, i.e. during and
0
200
400
600
0
200
400
600
after the scoliosis surgery, respectively. The images of rods were reconstructed in three-dimensions using
T5
quintic polynomial functions.[3] The intraoperative and postoperative corrective
forcesT5 were calculated
T6
T6
using the method based on Finite Element Analysis.[4]
T7
T7
The implant rod at the concave side had higher deformation than the convex side. The implant rods
T8
T8
had higher amount of deformation during the intraoperative than the postoperative phase. The results
T9
T9
indicate that the intraoperative corrective forces (below, right) were significantly higher than the
T10
T10
postoperative forces. The magnitudes of corrective forces imply that the intraoperative phase requires
T11
T11
more attention for preventing implant rod breakage or screw pullout during scoliosis surgery. Careful
T12
T12
intraoperative maneuver and planning is important to achieve a safe and optimal clinical outcome.
L1
After Surgery
L1
0
Patient
Patient3 1
Fixation
Level
FixationLevel
Before Surgery
600
0
T5
T5
T5
T5
T6
T6
T6
T6
T7
T7
T7
T7
T8
T8
T8
T8
T9
T9
T9
T9
T10
T10
T10
T10
T11
T11
T11
T11
T12
T12
T12
T12
L1
L1
L1
L1
00
200
200
400
400
600
600
00
Convex Side
200
400
600
200
200
600
600
400
400
T5
Corrective Force (N)
T5
Corrective Force (N)
T6
Intraoperative
T6
Postoperative
T7
T7
[2] Salmingo, R., et al. (2013) The Spine Journal (accepted, in press).
T8
T8
[4] Salmingo, R., et al. (2012)
Clinical Biomechanics
27(6):545–550.
Patient 2
Fixation Level
References
[1] Salmingo, R., et al. (2013) Clinical Biomechanics 28(2):122-128.
[3] Salmingo, R., et al. (2012) ASME J Biomech Eng 134:054502.
Concave Side
200
400
T9
T9
T10
T10
T11
T11
T12
T12
L1
L1
Multi-View Stereo Approach to Gait Analysis
Martin S. Christiansen1,2, Henrik Aanæs3, Thomas B. Moeslund4, Henrik Koblauch2, Tine Alkjær2, Erik B.
Simonsen2
1
Department of Research and Development, Danish Institute of Fire and Security Technology, Hvidovre,
Denmark
2
Department of Neuroscience and Pharmacology, University of Copenhagen, Copenhagen, Denmark
3
Informatics, DTU, Kgs. Lyngby, Denmark
4
Department of Architecture, Design and Media Technology, Aalborg University, Aalborg, Denmark
Introduction. Markerless motion capture has been studied intensively during the past decades [1-5], because
the usage of markers have several drawbacks. The placement of markers are time consuming, it requires
knowledge of human anatomy and small variations in the placement of the markers can induce large variations
in crosstalk between joint angles [6]. In addition, skin movement artifacts cause problems in estimating joint
angles with relatively small dynamic ranges such as the internal/external rotations [7].
Visual Hull (VH) models are commonly applied for 3D reconstructions in markerless motion capture. However,
VH models tend to overestimate the volume of the 3D reconstruction, because it cannot carve concave
structures in the 3D reconstruction. The overestimation of the 3D volume makes the thigh and shank axial
symmetric, which inhibits accurate estimation of the internal/external rotations [8]. The incapability of
estimating internal and external rotations also makes markerless systems infeasible for 3D inverse dynamics.
Multi-view stereo algorithms are capable of gaining highly detailed 3D reconstructions [9] and have been
applied to markerless motion capture recently [10, 11]. However, these approaches were not combined with
pose estimation and were therefore not applicable for analyzing joint angles.
We investigated the performance of joint angle estimation using high detailed 3D reconstructions rather than
the VH models in conjunction with model based motion capture. No errors of the internal/external rotations
have previously been qualitatively reported for markerless motion capture systems.
The present approach was compared with a marker based method. It is noted that marker based systems do
not provide ground truth data, and that are somewhat imprecise, e.g. due to skin movement artifacts (SMA)
and issues with placing the markers at well-defined anatomical positions, c.f. [7, 12, 13]. The results of this
study will therefore reflect the comparability to the marker based methods that are widely used and not
present the actual errors associated to the proposed markerless system. However, we believe the resulting
differences between the two systems are highly correlated with the ground truth error.
Methods. A 3D reconstruction was synthesized for each frame using the PMVS2 algorithm proposed by
Furukawa and Ponce [14] in combination with Poisson reconstruction.
An articulated model with 6 degrees of freedom for each limb segment was fitted to the 3D reconstructions
through the Iterative Closes Point Algorithm.
Data of a walking test subject was evaluated.
Results and conclusion. The markerless system was comparable to the results from marker based system.
However, we experienced difficulties in properly estimating internal/external rotations as a consequence of
gaps in the 3D reconstruction provided by limited insight of the medial side of the thigh from the camera
positions. More studies have to be made in dealing with these holes in order to obtain robust results.
References
[1]
[2]
[3]
[4]
[5]
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
D. Weinland, R. Ronfard, and E. Boyer, "A survey of vision-based methods for action representation, segmentation and
recognition," Computer Vision and Image Understanding, vol. 115, pp. 224-241, 2011.
M. B. Holte, C. Tran, M. M. Trivedi, and T. B. Moeslund, "Human Pose Estimation and Activity Recognition From Multi-View
Videos: Comparative Explorations of Recent Developments," IEEE Journal of Selected Topics in Signal Processing, vol. 6, pp.
538-552, 2012.
T. B. Moeslund, A. Hilton, V. Krüger, and L. Sigal, Visual Analysis of Humans: Looking at People: Springer, 2011.
R. Poppe, "Vision-based human motion analysis: An overview," Computer Vision and Image Understanding, vol. 108, pp. 4-18,
2007.
J. Xiaofei and L. Honghai, "Advances in View-Invariant Human Motion Analysis: A Review," IEEE Transactions on Systems, Man,
and Cybernetics, Part C (Applications and Reviews), vol. 40, pp. 13-24, 2010.
S. J. Piazza and P. R. Cavanagh, "Measurement of the screw-home motion of the knee is sensitive to errors in axis alignment,"
Journal of Biomechanics, vol. 33, pp. 1029-1034, 2000.
M. Akbarshahi, A. G. Schache, J. W. Fernandez, R. Baker, S. Banks, and M. G. Pandy, "Non-invasive assessment of soft-tissue
artifact and its effect on knee joint kinematics during functional activity," Journal of Biomechanics, vol. 43, pp. 1292-1301,
2010.
S. Corazza, L. Mündermann, A. M. Chaudhari, T. Demattio, C. Cobelli, and T. P. Andriacchi, "A Markerless Motion Capture
System to Study Musculoskeletal Biomechanics: Visual Hull and Simulated Annealing Approach," Annals of Biomedical
Engineering, vol. 34, pp. 1019-1029, 2006.
S. M. Seitz, B. Curless, J. Diebel, D. Scharstein, and R. Szeliski, "A comparison and evaluation of multi-view stereo
reconstruction algorithms," Proceedings of the IEEE Computer Society Conference on Computer Vision and Pattern
Recognition, vol. 1, pp. 519-526, 2006.
Y. Furukawa and J. Ponce, "Dense 3D motion capture for human faces," IEEE Conference on Computer Vision & Pattern
Recognition (CVPR), pp. 1674-1681, 2009.
L. Kun, D. Qionghai, and X. Wenli, "Markerless Shape and Motion Capture From Multiview Video Sequences," IEEE
Transactions on Circuits and Systems for Video Technology, vol. 21, pp. 320-334, 2011.
D. L. Benoit, D. K. Ramsey, M. Lamontagne, L. Xu, P. Wretenberg, and P. Renström, "Effect of skin movement artifact on knee
kinematics during gait and cutting motions measured in vivo," Gait & Posture, vol. 24, pp. 152-164, 2006.
L. Lucchetti, A. Cappozzo, A. Cappello, and U. D. Croce, "Assessment of non-optimal, least squares, and artefact-compensated
joint kinematics estimators," Journal of Biomechanics, vol. 31, 1998.
Y. Furukawa and J. Ponce, "Accurate, Dense, and Robust Multiview Stereopsis," IEEE Transactions on Pattern Analysis and
Machine Intelligence, vol. 32, pp. 1362-1376, 2010.
Title
Temporal structure of lower limb kinematics during walking at different velocities
Authors
Pet er C. Raffalt, Martin Kjær Guul, Andreas Nexmann Nielsen, Erik B. Simonsen and Tine Alkjær
Affiliations
Department of Neuroscience and Pharmacology
University of Copenhagen
Abstract
Variability is inherent in human walking and is related to the time-dependent structure of the lower limb
movements. It is unknown how this temporal structure changes with velocities above and below the
theoretical optimal velocity (here defined as 42 % of Froude speed). The present study tested the
hypothesis that a U-shaped relationship between the temporal structure of the gait kinematics and the
walking velocity would exist with a minimum at the theoretical optimal walking velocity. Twelve healthy
male subjects walked at their optimal velocity and 20, 40, 160 and 180 % of the optimal walking velocity for
12 minutes. Knee and ankle joint angles were measured using two electro goniometers (sampling
frequency = 333 Hz) placed on the lateral side of the left knee and ankle. The angle signals was down
sampled to 50 Hz and the largest Lyapunovs exponent (LyE) was calculated using Wolf’s algorithm. A Ushaped curve in LyE for the knee joint angle with a significant lower value at the optimal walking velocity
indicating higher local stability in the knee joint motion related to the theoretical optimal velocity and
partly confirming the hypothesis of the study. No differences in the ankle joint were observed.
The influence of gait speed on local dynamic stability and variability
during walking in transfemoral amputees. A nonlinear analysis
Magdalena Rós Guðnadóttir, Erik B. Simonsen, Tine Alkjær
Department of Neuroscience and Pharmacology
University of Copenhagen
Background: The concept of stability has proven to be difficult to define. Nonlinear methods have in the
past years been used to evaluate stability and variability under different conditions and for different patient
groups, like amputees. Contradictory results have been reported regarding the stability of walking and
whether slow walking is more stable than fast or the other way around. The purpose of the present study is
to answer the questions; Do amputees show less local dynamic stability and more variability in their
acceleration time-series than healthy controls? Does walking at higher speeds result in less local dynamic
stability and higher variability than walking at lower speeds? Do amputees show more muscle intensity
than healthy controls during walking? Method: Nonlinear methods were used to assess the differences
between transfemoral amputees and healthy controls during walking on a motorized treadmill at three
different speeds. Short- and long-term maximum Lyapunov exponents were used to evaluate the local
dynamic stability of three dimensional accelerations of the pelvis and the approximate entropy was used to
evaluate the variability of those accelerations. Furthermore, the differences between the two groups in
EMG intensity of four muscles of the leg were assessed. Five transfemoral amputees and five age and
gender matching control persons participated in the study. Results: In general, subjects had higher values of
local dynamic stability and approximate entropy at higher speeds, indicating less stability and more
variability. Amputees had higher values of the short-term Lyapunov exponent in most cases, indicating
more instability. Regarding the long-term Lyapunov exponent, there was no evident pattern indicating less
stability in either of the groups. Amputees had higher values of approximate entropy in majority of the
cases and showed significantly higher muscle intensity than the control group during walking at all
speeds.Conclusion: No decisive pattern was found regarding the differences in stability and variability
between amputees and healthy controls. However, there were some indications that amputees had less
local dynamic stability and more variability in their three dimensional accelerations. Amputees had
significantly greater muscle activity during walking at all speeds. This study also suggests that slow walking
is more stable than fast walking.
oral sessions
APPLICATION OF CLUSTER BASED EXPOSURE VARIATION
ANALYSIS IN BIOMECHANICS AND SPORT SCIENCE
1
2
1
Afshin SAMANI , Svend Erik MATHIASSEN , Pascal MADELEINE
1.
Center for Sensory-Motor Interaction (SMI), Dept. of Health Science and Technology, Aalborg
University, Aalborg, Denmark, [email protected] , [email protected]
2.
Centre for Musculoskeletal Research, Department of Occupational and Public Health Sciences,
Faculty of Health and Occupational Studies, University of Gävle, Sweden,
[email protected]
Corresponding author e-mail: [email protected]
Abstract
Purposes
We compared the efficiency of a conventional exposure variation analysis (EVA) and a new clusterbased method (C-EVA) in discriminating different exposure time patterns.
Methods
To provide a basis for the study, we simulated the extent of exposure variation in a cyclic movement
by two sets of parameters representing “small” and “large” exposure variation along each of its
dimensions, i.e., range, frequency and similarity. The simulation design rendered a
design of
exposure groups representing different sizes of the range (“far”, “near”), the velocity (“high”, “low”) and
the similarity of exposure sequences (“large”, “small”). For each exposure group, 120 simulations
termed “exposure trace” were performed. Each exposure trace was composed of 100 sequential
cycles.
C-EVA and conventional EVA were performed on exposure traces. C-EVA is a cluster based version
of EVA and instead of intervals of EVA categories, samples of exposure traces are assigned to their
closest pre-specified optimal clusters. The optimal locations of cluster centers are determined by a
combination of clustering techniques, e.g., K-means and clustering optimization, e.g., the gap analysis.
The marginal distributions of both EVA and C-EVA with respect to the level and frequency were then
computed by adding up cell values along both dimensions. Using the marginal distributions, a linear
classifier was applied to discriminate between the simulated exposure patterns. The C-EVA approach
was also applied to electromyographic activity of trapezius (ascending and descending parts), deltoid
anterior and serratus anterior during a standardized computer work before and after repeated bouts of
eccentric exercises over two consecutive weeks in the shoulder region inducing soreness.
Results
The accuracy of the classification was 47% and 52% for EVA and C-EVA, respectively. Both methods
failed to discriminate exposure patterns differing with respect to similarity. The effects of two bouts of
eccentric exercise were investigated and an adaptation, i.e. an increase in the proportion of low level
exposure, was captured by the C-EVA.
Conclusions
The C-EVA has specific properties such as the optimality of data reduction and the capability of
handling multiple exposure time lines in one comprehensive analysis. The C-EVA was slightly better
than the conventional EVA in discriminating simulated exposure traces known to differ in variation, but
both methods failed to detect differences in similarity. The experiences from performing C-EVA on
empirical EMG data demonstrated that the approach detects potentially important changes in
muscular activity following induced muscle soreness.
Keyword : Clustering, Statsitical simulations, Gap analysis
October 25th | Copenhagen
×

Report this document