THE BAMOS PROJECT

Biomechanical Analysis of Motor Development in Walking
by Ann Hallemans, Dr. Peter Aerts
Laboratory for Functional Morphology, University of Antwerp, Belgium

The Functional Morphology Lab was started in the early seventies by Prof. Dr. Frits De Vree, who retired last year. Over the years, the lab and its members gained a lot of experience in the field of biomechanics. Researchers covered a wide range of topics, concentrating especially on animal feeding (bat, gecko) and animal locomotion (lizard, eel, frog, galago, magpie) In recent years, attention has also been paid to bipedalism and quadrupedalism in apes (gibbon and bonobo). The focus of all research is to investigate how the morphological structure of an animal is adapted to perform a certain functional task.

The BAMOS project commenced as a PhD study. It is the first to consider human locomotion. However, for years researchers from our lab have been (and are still) focusing on the phylogenetic development of bipedal gait in different hominid taxa. Therefore the link to ontogenetic development of walking was obvious. Although the rest of the article will focus on the goals and achievements of the BAMOS project, it is not the only project in which the Vicon system is used in our lab. Some research has been done on locomotion of lizards using the system. At the moment a project on feeding in large lizards also makes use of the system.

BAMOS is short for "Biomechanische Analyse van de Motorische Ontwikkeling van Stappen" or in English "Biomechanical Analysis of Motor Development of Walking". The goal of the project is to go beyond a mere description of the gait pattern of children. We hope to give insight into the biomechanical processes that underlie developmental changes in motor skills in young children.

The development of a stable habitual walking pattern can be divided into two phases: the first phase spans the first three to six months after the child has performed the initial independent steps, and is characterized by a rapid change in all gait parameters. The second phase continues until the age of eight and is characterized by a further slow maturation of the gait parameters.

The focus lies on the first rapid developmental phase: the first five months of independent walking are taken into consideration. It comprises a cross-sectional study and a longitudinal follow-up study. In the cross-sectional study, we try to identify different walking strategies observed in new walkers. The aim of the longitudinal follow up is to determine an evolution (maturation) in the different observed walking patterns. As an extension of the BAMOS project, two undergraduate students are looking at differences between assisted walking and independent walking in toddlers.

Previous researchers paid special attention to changes in spatio-temporal characteristics in new walkers (Scrutton, 1969; Clark & Phillips, 1987, e.a.). They showed that small children walk with short quick steps (high cadence), and that they have an overall slow walking speed and walk with a broad base of support. Another important difference is that they spend more time with their two feet on the ground than adults. Their double support phase accounts for 35% of the gait cycle versus 20% in adults. This prolonged double support phase is thought to be an indicator for balance problems in toddlers.

The Helen-Hayes marker set-up

Some interesting studies on joint kinematics also exist (Statham & Murray, 1971, Sutherland, 1980; Grimshaw et al., 1998). Striking differences with the adult gait pattern are: the guard position of the arms (the arms are held high in flexion and abduction), simultaneous flexion of the hip and knee during the swing phase and flexion of hip and knee during the stance phase (a bent-hip-bent-knee walking pattern). A heel strike, as observed in mature walking, is not systematically present. Balance problems and immature co-ordination have been suggested as limiting factors in the development of walking. However, studies trying to link changes in balance control or co-ordination to observed changes in spatiotemporal or kinematic features are scarce. In this study the focus is on changes in balance control, co-ordination of movement and musculoskeletal control during this period.

A data collection session in progress.
Lots of toys and the presence of mum make it easier to make the children walk.

The experimental set-up consists of a large wooden walkway (3m long by 1.5m wide) surrounded by six infrared Vicon cameras (Mcam 460, 250 Hz.). Two force platforms (AMTI, 0.5m x 0.42m, 250 Hz.) are built into the walkway. The force measurements are synchronized with the video recordings. To measure pressure distribution under the feet, two pressure pads are placed on top of the force plates (Footscan Int., 0.5m x 0.4m, 250 Hz., 3.5 sensors/cm). This allows for scaling of the pressure data to the simultaneously recorded forces and greatly improves the accuracy of the pressure measurements. Since the pressure pads are very stiff, the force data are not affected.

The Helen-Hayes marker set-up is used for measuring full body kinematics. To prevent problems with marker plucking, markers are attached to a tight fitting suit. Foot markers are attached to socks or soft leather shoes.

Morphometrical analysis is performed to obtain individual information about segmental mass distribution for each child. Therefore, a 15-segment body model is used (Otten et al., in prep). Mass distribution is calculated from body measurements (7 measurements/segment: the length, depth and width at a proximal point, depth and width at broadest point, depth and width at a distal point) taken from photographs. We use photographs since we discovered that taking body measurements with calipers in these young children is too frightening for them.

A brainstorm session with Peter Aerts and Ann Hallemans (photo: B. Otten)

As mentioned earlier, we want to find out how balance control, co-ordination and control of movement are integrated to give rise to a "smooth" walking pattern. We also try to determine the consequences of immature co-ordination and control on the resulting walking pattern. Therefore, we perform a complete biomechanical analysis of the gait pattern of toddlers.
Ground reaction force patterns, body centre of mass movements and joint kinematics are described. However, we go beyond a mere description, trying to link the different observations to one another. Another important area of study is foot function in new walkers. We look at stability and roll-off patterns of the foot by considering the course of the centre of pressure under the feet. But also pressure distribution under different foot regions and relative loading of the foot are taken into consideration.

As mentioned earlier, we want to find out how balance control, co-ordination and control of movement are integrated to give rise to a "smooth" walking pattern. We also try to determine the consequences of immature co-ordination and control on the resulting walking pattern. Therefore, we perform a complete biomechanical analysis of the gait pattern of toddlers.

A father holds his child to take two photographs for morphometrical analysis.

Ground reaction force patterns, body centre of mass movements and joint kinematics are described. However, we go beyond a mere description, trying to link the different observations to one another. Another important area of study is foot function in new walkers. We look at stability and roll-off patterns of the foot by considering the course of the centre of pressure under the feet. But also pressure distribution under different foot regions and relative loading of the foot are taken into consideration.

In order to understand the driving forces of movement, inverse dynamical analysis is carried out. Linking joint kinematics to patterns of joint torques and power absorption/ generation curves brings insight into the control of movement on the level of musculoskeletal systems.

The most important feature of infant walking is the large amount of variability. This is a feature characteristic for a learning process. When considering foot function, three different roll-off patterns can be identified in toddlers. In the first pattern
contact is rapidly followed by foot flat. Then further roll-off of the foot is observed and push-off forces are generated under the hallux region. In the second pattern (flat foot walking), the entire plantar surface area immediately contacts the ground. In the third pattern (toe walking) initial toe contact is followed by foot flat (Hallemans et al.; 2003). This is completely different from mature walking, which is characterized by a clear heel strike followed by a gradual roll-off of the foot.

Also in ground reaction forces, three different patterns can be observed, being single hump, double hump and peak hump. Different ground reaction force patterns seem to coincide with differences in body centre of mass oscillations. In turn, these body centre of mass movements can be linked to joint kinematics (Hallemans et al.; in review).

Balance problems in toddlers, suggested by the prolonged double support phase and broad base of support, are confirmed by the large lateral oscillations of the body centre of mass, the large values of the medio-lateral ground reaction force component and the large plantar surface contact areas.The wobbling pattern of the centre of pressure under the feet of new walkers also points towards balance problems. It seems they are oscillating forward and backward while trying to move over the supporting foot. Despite this oscillation, the forefoot region carries the majority of the load. This is most likely because this region offers more muscular control to compensate for minor imbalances than the rearfoot or midfoot region. (Hallemans et al., 2003.)

When considering neural control of walking, there are some indications that simple monosynaptic reflexes could play an important role (at least during weight acceptance). This is not unlikely, since previous research on muscle activation (EMG studies) in young children already suggested that with the acquisition of a new motor skill, neural control shifts back to a more primitive level. It seems at foot contact, muscle force is inadequate to sustain body weight. Therefore the hip, knee and ankle flex, thereby leading to muscle stretch. Activation of a stretch reflex increases muscle tone, thus preventing further collapse of the leg. (Hallemans et al., in review)

We are currently concentrating on inverse dynamical analysis in order to gain further insight into the control processes that underlie the observed kinematic walking patterns in toddlers. In the future, based on the data from our longitudinal study, we hope to identify an evolution in the different observed walking patterns and their underlying control strategies. The ultimate goal is to give more insight into motor development and motor control in normal healthy children.

References:
Clark, J. & Phillips, S. (1987) The step cycle organisation of infant walkers J. Mot. Behavior 19(4), 421-433

Grimshaw et al (1998) The 3-dimensional kinematics of the walking gait cycle of children aged between 10 and 24 months: cross-sectional and repeated measurements Gait & Posture 7, 7-15

Hallemans, A.; D'Août, C.; De Clercq, D.; Aerts, P. (2003) Pressure Distribution Patterns under the Feet of New Walkers: The First Two Months of Independent Walking Foot & Ankle Int. 24(5), 444-453

Hallemans, A.; Otten, E.; De Clercq, D.; Aerts, P. (in review) Immature control of walking increases variability of ground reaction force patterns and body center of mass movements in toddlers

Scrutton (1969) Foot print sequences of normal children under five years old Dev Med Child Neurol 11, 44-53

Statham, M. & Murray, M. (1971) Early walking patterns of normal children Clinical Orthopedics 79, 8-24

Sutherland, D.; Ohlsen, R., Cooper, L.; Woo, S. (1980) The development of mature gait J. Bone Jnt Surgery 62-A(3), 336-353