13:30   Musculoskeletal Systems 2: Postural Control
Chair: Jaap Harlaar
13:30
15 mins
THE EFFECTS OF DECOUPLED PELVIS ROTATION ON BALANCE CONTROL IN SITTING
Paul van Geffen, Jasper Reenalda, Peter Veltink, Bart Koopman
Abstract: Many wheelchair users with reduced postural stability have problems with functional movement during activities of daily life. To increase the base of support and to bring the body centre of mass closer the support surface, a kyphotic spinal posture with posterior tilted pelvis is often adopted. Because the pelvis forms the basis for trunk support and directly affects lumbar spine curvature, sagittal pelvis alignment plays an important role in balance control during trunk and upper extremity movement. Interventions that adjust pelvis angle in sagittal direction might therefore be applicable to regulate spinal posture and to improve functional performance in daily wheelchair use. The present study uses a simulator chair that rotates the pelvis independent from the trunk and seat support, and investigates the effect of so-called decoupled pelvis rotation (DPR) on balance control in ten subjects with a spinal cord injury (SCI). Experiments were performed with a computer-aided simulator chair instrumented with a parallelogram design that adjusts the backrest in sagittal direction and decouples the alignment of the pelvis from the trunk and seat support. The experiment involved four trials, each investigating the effects of chair modification on the displacement of the hands and centre of pressure (cp) during a controlled bimanual forward reaching task. Six uni-axial load cells (FUTEK®, California, USA) are mounted under the seat support for accurate measurement of cp. An infrared camera motion capturing system (VICON®, Oxford, UK) was used to capture hand movement from reflective markers that were placed on the third metacarpal joints. Preliminary results show that for all subjects the cp shifts in forward direction when the pelvis rotates anteriorly due to forward adjustment of the backrest. It was found that for high SCI-subjects the absolute and relative displacement of the cp and hands during maximal forward reach were smaller in posterior tilted pelvis postures. For low SCI-subjects on the other hand, relative cp and hand displacements was larger in posterior tilted pelvis postures, but no differences were found in absolute forward displacements. Although minimal effects were found in reaching distance for low SCI-subjects, it was observed that more effort is needed to reach forward in posterior tilted pelvis postures. In high SCI-subjects, larger displacements of the hands and cp suggest that anterior pelvic tilt benefits the ability to reach forward and is therefore associated with an improved balance control in sitting. Although the influence on reaching distance seemed highly dependent on the users’ functional capacity, we conclude that DPR affects balance control in SCI-subjects and that anterior pelvis rotation potentially benefits the users functional performance in daily wheelchair use.
13:45
15 mins
EVALUATION OF AN AMBULATORY MEASUREMENT SYSTEM FOR GAIT STABILITY
Sjoerd Bruijn, Jaap van Dieen, Warner Ten Kate, Onno Meijer, Peter Beek
Abstract: Falls pose a large problem in modern society [1]. Consequently, there is a growing body of research focusing on stability of posture and gait. Recently, two “dynamical systems” methods for quantifying stability have been applied to human gait; local dynamic stability, expressed as maximum Lyapunov exponents [2], and orbital stability, expressed as Floquet multipliers [3]. These measures appear to differentiate between age groups, and between patient and control groups [2-4]. In most literature, data from optoelectronic measurement systems were used to calculate these measures. Using wireless inertial sensors may be more practical when using these measures in a clinical, or home-based setting. This application of inertial sensing is highly feasible because in principle these stability measures do not require the alignment of the data to a global coordinate system. Although this approach holds promise, a direct comparison between estimates of Lyapunov exponents and Floquet multipliers obtained from optoelectronic and non-aligned inertial sensor gait data has up to date not been reported. We measured gait kinematics of nine healthy male volunteers, walking 5 minutes at three different velocities, using clusters of 3 infrared LED’s (used for movement registration with an Optotrak system(NDI, Canada)), fixed with neoprene bands to the right calf, right femur, the pelvis and the thorax. Under the bands, wireless inertial sensors [5] (which include 3D gyroscopes and accelerometers) were placed. Sample rate was 50 Hz. A 12 D state space was reconstructed for both the inertial sensor and the Optotrak data, using the 3D angular velocities and 3D linear accelerations and their time delayed copies. From these state spaces, maximum Lyapunov exponents were estimated as the slope of the average logarithmic divergence from two nearest neighbours [6]. Floquet multipliers [3], were calculated as the largest Eigenvalue of the Jacobian mapping S*-Sk to S*-Sk+1 for each phase in the stride cycle, where S* is the average trajectory in state space at a certain phase in the stride cycle, and Sk represents the values of the trajectory in state space at stride k during the same phase. The non-aligned inertial sensor data generally yielded stability measures that were highly correlated with results obtained from optoelectronic position data. It should be kept in mind that the relationship between these measures and a person’s resilience to a perturbation in real-life is yet to be established. Nonetheless, a recent study [4] showed that Floquet multipliers discriminated a group of elderly with a history of falling from a group of elderly without such a history.
14:00
15 mins
GLENOHUMERAL STABILITY DETERMINED FROM MUSCLE ACTIVITY: FROM MODEL SIMULATION TOWARDS IN VIVO MEASUREMENTS
Frans Steenbrink, Carel Meskers, Rob Nelissen, Jurriaan de Groot
Abstract: In the shoulder there is a delicate mechanical balance between glenohumeral stability and arm mobility.1,2 In case of massive cuff tears compensation of the lost cuff muscles elevation moment by the deltoids induces an increased upwards directed force component on the humeral head jeopardizing glenohumeral stability.3 We hypothesize that glenohumeral stability can be preserved at the cost of arm mobility by adductor muscle co-contraction during arm elevation tasks1,3, but a causal relation between compensating muscle activity and glenohumeral (in)stability was not yet proved. We investigated this relation using biomechanical model simulation (Delft Shoulder Model5) to study the effect of applied forces at different moment arms on muscle activity and glenohumeral (in)stability, preceding an in vivo experimental testing protocol. In a simulated cuff tear condition (fully torn supra-and infraspinatus) external downward directed forces, requiring arm elevation moments, were applied at the humeral shaft. Increasing moment(arms) at constant forces, as well as constant moments, by means of increasing moment arms and matching decreasing forces, were simulated. Deltoid forces are expected to increase with increasing momentarms at constant forces, affecting glenohumeral stability and causing adductor muscle co-activation. Under constant moment condition, larger moment arms coincide with decreased external forces and therefore deltoid forces and glenohumeral stability are not expected to be affected dramatically. Increasing moment arms at constant forces resulted in increased estimated deltoid activation, inducing glenohumeral instability and introducing adductor muscle co-activation (teres minor) to stabilize the glenohumeral joint. Glenohumeral stability appeared to be rather insensitive to external force magnitude. Stability was preserved at low moment arms irrespective of the external force magnitude (max Fext=200N), which made co-activation superfluous. When instability was induced, at higher moment arms, even the smallest applied external force in our simulations (Fext=25N) engenders glenohumeral instability. A constant moment applied at a small and large moment arm, with respectively large and small external forces, did not affect muscle activity or glenohumeral stability. The challenge now is to prove this concept in vivo. A special developed measurement device using a cardan axis with adjustable arm lengths will allow us to apply external forces at large and short moment arms in patients suffering massive rotator cuff tears, similar to the described simulations. This experiment is expected to demonstrate a causal relation between glenohumeral stability and compensating muscle activity of the deltoids and the arm adductors. In contrast to external force application relatively far from the glenohumeral joint, external forces close to the glenohumeral joint are not expected to affect glenohumeral stability. This is expected to result in relative lower abductor (deltoids) muscle activation, monitored by EMG, without endangering glenohumeral stability. Therefore no adductor muscle co-activation will be required. REFERENCES [1] J.H. de Groot et al., Clin.Biomech. 21, 2006. [2] S.P. McCully et al., J.Biomech. 40, 2006. [3] F. Steenbrink et al., Man.Ther. 11, 2006. [4] F.C.T. van der Helm, J.Biomech. 27, 1994. [5] H.E.J. Veeger and F.C.T. van der Helm, J.Biomech. 40, 2007.
14:15
15 mins
CONTROLLED ROTATIONAL MOTION EMPLOYED AS A MATHEMATICAL BENCHMARK STANDARD FOR THE RESPONSE OF BI- AND TRI-AXIAL ACCELERATION SENSORS
Hans Sauren, Stan Laudij, Luc de Witte
Abstract: As falls and fall-related injuries remain a major challenge in the public health domain, reliable and immediate detection of falls is important so that adequate medical support can be delivered. Commonly used sensors in fall detection are generally based on measuring accelerations of the human body in two or three carthesian or perpendicular dimensions (bi- or tri-axial accelerometers). Human gait patterns recorded with acceleration sensors show highly complex and irregular spiky structures; to discriminate unequivocally between falls and Activities of Daily Living (ADL) is frequently an arduous and successfully limited enterprise. As a possible way to enhance comprehension of the fall information partly obscurd in human gait patterns and concomitantly, to circumvent the difficulties associated with logging and data interpretation, a preliminary study was conducted during which the responsivity of a bi-axial accelerometer was investigated by testing its capability and accuracy of tracking mathematically defined trajectories. The advantage of this tracking method is obvious: the pattern of movement is mathematically unique and apriori known in all its details. The predefined mathematical trajectory serves as a standard of movement in an absolute sense. Therefore, the responsivity of an acceleration sensor can be unequivocally compared to the mathematical standard. As a further advantage, the robustness and sustainability of algorithms commonly used in gait pattern analysis and data processing become accessible to further mathematical scrutiny. During fall the human body may reach accelerations of the order of two (2 g) or three (3 g) times that of a free falling body in the gravitational field of the earth. To achieve accelerations of this order in a laboratory, circular movements with varying velocity functions were employed. To experimentally meet this goal, a bi-axial accelerometer was rigidly mounted on the rim of circular disc, the diameter of which was about 30 cm. The rotational movements of the disc were controlled by a voltage controlled power supply. Different modes of movement, such as sinusoids and square-waves, were selectable from a software programme written in LabView\texttrademark. The polar movements resolved into the rectangular coordinates of the acceleration sensor yield after some basic vectorial and algebraic calculations: \begin{eqnarray*} a_{x} & = & \ddot{x}(t) = -\dfrac{v^{2}(t)}{\varrho} \cos\big(s(t)\big) -\dot{v}(t)\sin \big(s(t)\big) \\ a_{y} & = & \ddot{y}(t) = \dot{v}(t)\cos\big(s(t)\big)-\dfrac{v^{2}(t)}{\varrho}\sin \big(s(t)\big) \end{eqnarray*} in which $s(t)=\int v(t)\textrm{d}t$ and $\varrho$ is the measure of curvature of the chosen trajectory. %It is mathematically trivial to show that the magnitude of the vector of acceleration is %\begin{equation*} %|\vec{a}| = a = \sqrt{a_{x}^{2}+a_{y}^{2}} = \sqrt{\Big(\dfrac{\textrm{d}v}{\textrm{d}t}\Big)^{2}+\dfrac{v^{4}}{\varrho^{2}}} %\end{equation*} The transformed equations above make it feasible to substitute any kind of time-dependent function of velocity $v(t)$. Data acquired by employing voltage fuctions differing mathematically in shape, such as uniform circular motion $v(t)=$ const., sinusoids $v(t)=\sin \omega t$, and square-wave functions are currently being processed, the results of which will be published in due time. Falls of humans may be simulated by using square-wave velocity functions with very short periods or equivalently, using high frequencies. In this way the complexity of the human gait pattern may be simulated and may become available to mathematical analysis in a controlled way. Future research in the field of mathematically controlled movements is directed towards application of more advanced and sophisticated movements in three dimensions. To achieve these goals, it is intended to mount tri-axial accelerometers to robots, such as the Haptic Master, capable of performing spatially controlled movements.
14:30
15 mins
A REALTIME MONITOR FOR FALLS AND STUMBLES DURING ACTIVITIES OF DAILY LIFE
Kenneth Meijer, Rachel Senden, Emile Arnoldussen, Bernd Grimm, Ide Heyligers, Laurence Kenney, Stephen Preece, Hans Savelberg
Abstract: Falling is one of the leading causes of hospitalization in people over 65 years of age. Slips and trips are the leading causes of falling incidents [1]. The increased risk of falling in the elderly could be attributed to i) a higher probability of stumbling due to poor obstacle avoidance resulting from such factors as reduced foot clearance, vision or attentional/cognitive capacity, or increased reaction times ii) a poorer recovery strategy once a stumble has occurred [2]. It is currently unknown if elderly experience more stumbling incidents during activities of daily life (ADL) than younger subjects, because there are no reliable means to obtain the relevant data. The aim of the present study was to develop and evaluate a real time stumble and falls detector, which can identify stumbling and falling incidents during regular daily activities. A wearable stumble and fall detector was developed based on PASAQ technology [3]. The device consisted of a three axial accelerometer, an ECG front-end, a microprocessor and a Bluetooth™ module to communicate stumbling and falling events to a mobile handset. The unit was programmed with algorithms to distinguish dynamic (i.e. locomotion) and static (i.e. sitting, lying & standing) activities and to detect stumbling and falling events during dynamic activities. The device was evaluated in two small user trials; i) an experiment in which subjects walked on a treadmill at different speeds while stumbles were induced via a rope attached at the ankle plus a 20 minute long standardized activity protocol in the laboratory (2 elderly & 3 young) and ii) a one day measurement in a retirement home (n=6). All subjects signed an informed consent and the experiments were approved by the relevant medical ethical committees. The first set experiments showed that the sensitivity of the device was reasonable (40-80%) and that stair descent and rapid transfers between activities during the standardized activity trial resulted in falsely detected stumbles, particularly in the young adults. During evaluation period of the trials (10 +/- 3.9 hours) at the retirement home none of the subjects experienced any stumbles or falls. However, the device detected several stumbles in these subjects (2-22 counts (min-max)) during the performance of daily activities. To our knowledge this is the first study that has developed a device that has the potential to measure precursors of falling incidents, which may play an important role in the prediction of future falling events. Further validation studies and optimization of the device are currently being conducted. This study was made possible by a grant from Vodafone R&D, The Netherlands REFERENCES [1] Berg W.P., H.M. Alessio, E.M. Mills and C. Tong, Circumstances and consequences of falls in independent community-dwelling older adults, Age Ageing 26 (4) (1997), pp. 261–268. [2] Dieën H v., M. Pijnappels and M.F. Bobbert, Age-related intrinsic limitations in preventing a trip and regaining balance after a trip, Safety Sci 43 (2005), pp. 437–453. [3] www.maastrichtinstruments.com/files/PASAQ0-173-AA01%20Flyer%20PASAQ%20ENGELS.pdf