10:40   Musculoskeletal Systems 1: Modelling
Chair: Bart Koopman
15 mins
Iris Busscher, Albert van der Veen, Idsart Kingsma, Jaap van Dieen, Bart Verkerke, Albert Veldhuizen
Abstract: • Background The spine is a complex structure which provides stability and allows motion in flexion, extension, lateral bending and axial rotation. Different regions of the spine behave differently in motion and have their own dominant function. Quantitative information on this behaviour is scarce and can only be derived from biomechanical in vitro testing of spinal segments. The aim of this study is acquire quantitative information about the biomechanical properties of multi-level spinal segments: the range of motion (ROM), neutral zone (NZ), neutral zone stiffness (NZstiff) and flexibility of three thoracic and one lumbar region of the spine. • Materials and Methods 24 spinal segments of 6 human cadaveric spines were prepared for biomechanical testing. Each specimen contained 4 vertebrae and 3 intervertebral discs: T1-T4, T5-T8, T9-T12, L1-L4. Pure moments were applied to each specimen to a maximum of 4Nm in flexion/extension, lateral bending and axial rotation. Displacement of individual vertebra was measured using an optoelectronic 3D movement registration system. Biomechanical parameters (ROM, NZ, NZstiff, flexibility) of the spinal regions were calculated from the acquired load-displacement data (figure 1). • Results Tests showed significant differences of ROM between the regions in all directions. The NZ was significantly different between the regions in axial rotation and lateral bending, but not in flexion/ extension. A trend could be seen between size and stiffness of the neutral zone. A large neutral zone was associated with a high stiffness, a small neutral zone with a low stiffness. Flexibility of the spinal regions was variable; no significant differences were found between the four spinal regions. • Discussion This study showed the differences in biomechanical characteristics of different regions of the human spine. Understanding the biomechanics of the spine is crucial to advances in the understanding of spinal behaviour, clinical care, development of new treatment methods for spinal disorders and for validation of finite element models of the spine which are increasingly being used for pre-clinical evaluation of new implants and new surgical procedures. In contrary to other literature, multi-level segments of the whole thoracolumbar spine were tested. Therefore results are more in agreement with the normal physiological situation and characteristics of different regions are truly comparable. • Acknowledgements This study is financed by STW, the Dutch Technology Foundation
15 mins
Christiaan Jansen, Bart Verkerke, Jasper Homminga, Nico Verdonschot
Abstract: Introduction: The prediction of the (time dependent) response of the spine to different mechanical loading conditions is essential in understanding degeneration mechanisms leading to injuries, loss of spinal height, change in kinematics, vertebral fractures and treatment options. The spine is a complex chain of rather rigid and flexible elements. Most current FE models, however, focus on either the vertebrae or the intervertebral disks. We feel that this is a simplification of reality because the behaviour of the vertebrae is influenced by the intervertebral disk and visa versa. Our approach is therefore to develop a model that includes a realistic mechanical behaviour of both the disk and the vertebrae. Methods: This abstract reports on the development and validation of our nonlinear FE model that can quantify the mechanical responses of the L2-L4 motion segment to varying external loads. To this end, we obtained a radiographically normal cadaveric L2-L4 section from an eighty-one year old female. The geometry of this spine specimen was measured using CT-imaging. The mechanical behavior of this spine specimen was measured with in vitro tests that used Radiostereometric Analysis (RSA) to accurately measure its deflecting behavior over time. The CT-data as well as some literature data was then used to build the FE-model.In total the model needed nearly 600,000 tetrahedral elements to have a geometrical accurate model. The material properties of the 380,000 elements representing the vertebral bone were based on the CT-data by converting the calculated weighted calcium density value of each solid element. The two disks, each containing an annulus and nucleus were modelled with in total 190,000 tetrahedral elements while the disk fibres (giving the annulus anisotropic characteristics) were modelled with a few thousand nonlinear truss elements. The ligaments around the spine were also simulated using for about one-hundred nonlinear truss elements. Results and discussion: Six loading conditions (flexion and extension, left and right lateral bending, left and right torsions) were done during the experiments and demonstrated significantly different behaviour of the two disks. For extension the upper disk was found three times more stiff than the lower disk, while for flexion this ratio was found to be nine. Such behaviour indicates spinal degeneration. The results of the developed non-linear model showed a good agreement with our RSA experiments as well as with data found in literature. We are currently performing fracture tests on vertebrae to enable implementation of vertebral fracture in the model and implementing bone adaptation in the model. The detailed estimates of the state of stress/strain of this validated FE model can be used to test current medical solutions and can be a valuable tool in new developments for interventions on the degenerative spine.
15 mins
Natallia Dvinskikh, Leendert Blankevoort, Mahyar Foumani, Simon Strackee, Jos Spaan, Kees Grimbergen, Geert Streekstra
Abstract: Introduction. The wrist is a compact and complex human joint. Therefore the behavior and function of its components during motion are hard to study. A biomechanical joint model is one of the ways to understand the biomechanics of the wrist and its components better. The enhanced knowledge about wrist biomechanics may lead to improvement of clinical diagnoses and treatment planning. Purpose. The objectives of the presented study are: 1) to develop a biomechanical wrist model, based on cadaver specimen geometry and 2) to study the motion patterns and the functions of the wrist structures during motion. Materials and Methods. The geometry of the entire wrist was reconstructed from cryomicrotome images with a resolution of 0.1x0.1x0.1 mm3 [1]. The multi-body model of the wrist was based on the detected joint geometry. The model calculates the relative position and orientation of bodies connected by elastic links, like ligaments and articular contacts, for given external loads and kinematic constraints [2]. In the model parameter study, thickness and elastic modulus of the cartilage contacts as well as initial length and elasticity of the ligaments were varied during the simulations [3]. The simulations were separately performed for flexion/extension and radial/ulnar deviation of the wrist. Results. Intra- and inter-observer experiments reveal that the acquired wrist geometry is accurate and suitable for the development of a biomechanical wrist model. Over a certain range of wrist motion, in flexion-extension as well as ulnar-radial deviation, the constraint moment was almost negligible (free motion range). The cartilage contact surfaces guided the motion of the wrist within this range. Beyond the free range of motion, the constraint moment increased progressively as caused by the ligaments and articular contacts. Parameter variations of the cartilage properties and ligaments elasticity have a small impact on the model behavior while small changes in the initial length of the ligaments cause significant variations of the model response. Conclusions. The model simulations show that the extrinsic wrist ligaments function as limiters of joint motion. Due to the ligament forces at the extremes of motion and their interaction with the articular contact forces, mobility of the wrist is limited and the joint is restrained. Initial length of the ligaments is the most important model parameter. The precise reconstruction of the entire wrist geometry from the cryomicrotome images opens up new opportunities for biomechanical studies of the wrist with a realistic joint geometry. REFERENCES [1] J. A E Spaan et al. “Visualisation of intramural coronary vasculature by an imaging cryomicrotome suggests compartmentalisation of myocardial perfusion areas”, Med Biol Eng Comput, Vol 43, pp. 431-435, (2005). [2] S. D. Kwak, L. Blankevoort, and G. A. Ateshian. “A mathematical formulation for 3d quasi-static multibody models of diarthrodial joints”, Comput Methods Biomech Biomed Engin, Vol 3, pp. 41-64, (2000). [3] H.H. Savelberg et al. “Strains and forces in selected carpal ligaments during in vitro fexion and deviation movements of the hand”, J Orthop Res, Vol 10, pp. 901-910, (1992).
15 mins
Wiebe de Vries, Dirk Jan Veeger, Chris Baten, Frans van der Helm
Abstract: Main features: To generate input for a biomechanical model of the upper extremity from inertial magnetic sensors, local frames of reference for each segment need to be determined. The sensors deliver orientation estimations based on acceleration, angular velocity and magnetometer data, but no 3D position information. The option of building the local reference frames based upon functional axes of rotation is addressed. When performing a series of well-defined, isolated, uni-axial rotations of a segment, the angular velocity vectors can be used as an estimator for the orientation of the axis around which the segment rotates. With two of these vectors, e.g. thoracal flexion and torsion, a local frame of reference can be defined. Subjects were equipped with sensors (MT-X, Xsens, Netherlands) on thorax, humerus, forearm, and hand. An optoelectronic system (Vicon, 6 MX13 cameras) served as reference system. Reflective markers were attached to sensors and on Bony Landmarks conform the ISB proposal [1]. Prior to subject measurements, both systems were aligned, to enable all data to be expressed in the same global reference system. Subjects were asked to perform the following movements, five repetitions each: Trunk: flexion- retroflexion, lateral flexion, and torsion; Humerus: ab-adduction, flexion-anteflexion, endo-exo rotation; Forearm: elbow flexion-extension, pro-supination; Hand: ulnar-radial deviation, palmar-dorsal flexion. Results: The method was well repeatable, with acceptably small errors (up to 2 º standard deviations over 6 trials). Differences in orientation of local reference frames varied from 5 to 20º, over segments, but also over the two subjects. Conclusions: For most segments at least two functional axes can be found with a repeatability error of less than 2º. Comparison with a Vicon based bony landmark method showed differences in local coordinate systems of 5 to 20º. More subjects need to be measured to evaluate a possible ‘correction method’ for these differences. The determination of a scapular frame of reference with a sensor is not trivial; prescribed scapular movement is not under voluntary control; alignment with the thorax reference frame will lead to an unacceptable offset. Estimation of scapular orientation by means of regression equations is still the best alternative. REFERENCES [1] Ge Wu, F van der Helm, HEJ Veeger, et al, ISB recommendation on definitions of joint coordinate systems… Part II, Journal of Biomechanics 38 (2005) 981–992
15 mins
Peter Krekel, Frans Steenbrink, Rob Nelissen, Jurriaan de Groot
Abstract: Shoulder replacement surgery frequently leads to limited Range of Motion (RoM) of the glenohumeral joint. This limitation occurs actively, when RoM is limited by the impaired state of muscles, as well as passively, when RoM is limited by bony structures and shortened capsular-ligamentous structures.1 In order to improve the outcome of shoulder replacement surgery with regards to RoM, a CT-based pre-operative planning system was developed to predict the bone constrained RoM of the glenohumeral joint.2 Although these simulations assist the surgeon in avoiding bony impingement, prediction of active and passive RoM limitations imposed by muscles and ligaments is not incorporated in the simulator. Determination of the relationship between bone constrained, passive and active RoM contributes to the applicability of the RoM simulator and further assists the surgeon in setting up an optimal pre-operative plan. In a pilot study (n=1) to determine the contribution of passive structures to glenohumeral RoM limitations, we used the Flock of Birds electromagnetic tracking device to measure active and passive RoM of a healthy subject. The subject was instructed to position his arm in a plane of elevation of 0°, elevate to 45° and refrain from internal rotation. From this initial position, the subject was instructed to perform 16 systematic variations of ab-/adduction and retro-/anteflexion. In order to measure maximal active RoM the subject was instructed to move the arm as far as possible. Subsequently maximal passive RoM was recorded by moving the arm until considerable resistance was reached. The CT-scan and motion measurements were imported in the RoM simulator. Bone constrained RoM was calculated and compared to the measured active RoM and passive RoM. The results show that the bone-determined RoM for abduction is 60° larger than the RoM actively used. By passively moving the arm, an additional 18° can be reached. However, it can still be concluded that much of the RoM that is allowed by bony structures is unused as a result of other limitations. Our results also show that active, passive and bone-determined RoM are almost equal to each other for retroflexion and adduction. This indicates that these motions are barely affected by tissue limitations. This study demonstrates the relationship between the different types of RoM for a healthy shoulder. In future work a larger study will be conducted to investigate the variation of these findings. In addition, the RoM of patients will be evaluated to determine the relationship between active, passive and bone constrained RoM for various types of prostheses.