[Systematic analysis of ground reaction forces before and after hip and knee arthroplasty]. / Systematische Analyse der Anderung der Bodenreaktionskräfte bei Hüft- und Kniearthrose prä- und postoperativ.

In this prospective study we employed a newly developed gait analysis system to compare the ground reaction force patterns in 15 patients before and after total hip or knee replacement. In this system, data are measured separately for each limb. Measured data were also obtained from 30 healthy adults and compared with those obtained from the patient group. We analysed the three-dimensional force patterns, impulse, frequency, stride and double stance, and compared changes in the postoperative gait patterns. The vertical force maxima Fy identify the peak forces obtaining during walking. The results showed significantly increased (p < 0.05) postoperative force maxima Fy2 and Fy3 for both knee replacement (Fy2: 82.48 to 86.17 and Fy3: 96.09 to 99.35% body weight, pre- and postoperatively, respectively) and hip replacement (Fy2: 84.44 to 88.08 and Fy3: 98.63 to 101.96% body weight, pre- and postoperatively, respectively). The ADAL system proved suitable for the easy performance of gait analysis, and thus may be of future value in the area of clinical quality assurance.

[Bone adaptation changes mechanical stress in the femur--a prospective two years follow up after Hüft-TEP implantation]. / Knöcherne Anpassung ändert mechanische Beanspruchung des Femurs--eine zweijährige prospektive Studie nach Hüft-TEP Implantation.

AIM: Periprosthetic adaptive bone remodeling after total hip arthroplasty (THA) has been frequently simulated in computer models, combining bone remodeling theory with finite element analysis. Unfortunately, there is still a lack of clinical validation data. The collection of prospective volumetric bone density data with a clinical computerized tomography study after THA was previously conducted. The objective of the study presented here is an additional evaluation of femoral strain. METHOD: In a prospective clinical trial with 7 female patients after THA computerized tomography (CT) and finite element (FE) modeling was patient specifically combined immediately after surgery, as well as at three and 24 months postoperatively. Mechanical strain was expressed by strain energy density. RESULTS: Corresponding to a bone density decrease, a decrease of the femoral strain energy density was observed during the two year follow-up after surgery (approximately 20 %). The predominant change occurred within the first three months and was found to be distally stronger than proximally. CONCLUSION: As far as we are aware, this is the first combination of fully prospective 3D CT density data in vivo with patient-specific finite element modeling. The assessment of mechanical strain data during a follow-up trial should be a new approach for analyzing hip stems in clinical biomechanics. A future confirmative study based on a statistical case number calculation would increase the evidence of the results presented here.

Development of a hybrid finite element model for individual simulation of intertrochanteric osteotomies.

Intertrochanteric osteotomies of the proximal femur are used to improve the anatomy and function of the hip joint in a number of orthopaedic diseases. To investigate the geometrical and biomechanical aspects of pre-operative planning we created a set of programs to automatically perform a simulation of intertrochanteric osteotomies on a three-dimensional finite element model of the human proximal femur based on computed tomography (CT) data and using uniform brick-shaped elements. To eliminate artefacts resulting from the rough surface of the brick elements, the femoral head was represented by a tetrahedron-based head that included a cartilage layer and a subchondral cortical zone. Applicability of the procedure was tested by performing a parametric study using a model created from CT scans taken in vivo, by applying individually calculated force conditions for the one-leg stance situation. We found a large influence of osteotomy angle on the observed stress in the femoral head cartilage, especially in a situation with insufficient containment of the femoral head. The model presented here is a biomechanical tool to simulate intertrochanteric osteotomies patient-specifically for a better understanding of the effects of such operations in the individual case. The open design of the described programs allows future interfacing with surgical navigation and robot systems.

[Effects of shaft length of finger joint prostheses on tension distribution in the bone]. / Auswirkung der Schaftlänge von Fingergelenkprothesen auf die Spannungsverteilung im Knochen.

In order to investigate the influence of stem length in finger joint prostheses on stress in the surrounding bone area, finite element (FEM) calculations of finger bones before and after prosthetic replacement of metacarpophalangeal joints with cementless implants of different stem length were performed. CT scans of the metacarpal bone and proximal phalanx before and after implantation of a prototype of a noncemented semiconstrained implant for the MP joint, which has been developed to replace metacarpophalangeal joints destroyed by rheumatoid arthritis, were analysed. The FEM calculations showed comparatively decreased differences of the von-Mises stress after implantation of intramedullary stems reaching the middle of the diaphysis. At the metaphysis of the metacarpal head we found an increase of the von-Mises stress of 1.3 MPa (Mega Pascal = 10(6) Pa), an increase of 18.9 MPa around the shortest prosthesis and a decrease of 21.4 MPa around the prosthesis with the longest stem.

The long-term (8-12 years) results of valgus and lengthening osteotomy of the femoral neck.

Shortening of the femoral neck and proximal displacement of the greater trochanter are the principal complications following avascular necrosis of the capital epiphysis head in early childhood. We report here the long-term follow-up of a series of osteotomies performed to lengthen the femoral neck and thus to restore the normal anatomy and function of the hip joint. Out of a sample of 24 patients, 15 (62.5%) were reviewed at the end of a mean follow-up of 10 years and 2 months. Fourteen of them (93.3%) had originally complained of pain on walking which varied in severity. After the operation, five were free from pain at the end of the follow-up, 9 still experienced pain on walking, and 1 also reported pain at rest. Before the operation, a positive Trendelenburg's sign was found in 8 of the patients. This was still present in 2 at follow-up. The average perpendicular distance from the center of the femoral head to a horizontal plane passing through the tip of the greater trochanter (the centro-trochanteric distance) was reduced from 33 mm to 7 mm. Apart from correcting the anatomical deformity, it is suggested that this operation may well provide lasting relief from pain and increase the power of the abductor muscles. It may also delay the onset of osteoarthritic change and in this way postpone the necessity for an endoprosthesis.

[Influence of cartilage cap modelling on FEM simulation of femoral head stress]. / Einfluss der Knorpelkappenmodellierung auf die FEM-Simulation der Femurkopfbeanspruchung.

The present study reports on the finite element analysis (FEA) of the femoral head in a process of preparation for a program for the realistic simulation of correctional osteotomies of the proximal femur. While the material properties have been studied extensively, only few publications consider the influence of the cartilage layer geometry on FE stimulation of the hip joint. Various models of the femoral head with and without the cartilage layer are generated and analysed. On looking at the maximum surface stresses, we found a strong influence of the cartilage layer and the subchondral osseous layer on the magnitude of the von Mises equivalent stress. The model with an anatomically realistic cartilage layer and compact bone shows stresses of between 4 and 5.5 MPa, depending on the position of the joint, while the model with a concentric cartilage layer has a maximum von Mises stress of 0.8 MPa. Only on simulation of a "realistic" cartilage layer, with a maximum thickness at the "pole" and minimum thickness at the "equator" do the changes in stress distribution--determined by changes in the position of the femoral head--become visible. Owing to major artefacts and the inability to create a realistic cartilage layer, voxel-based models of the femur are not suitable for the simulation of the femoral head surface.

[Testing dental implants with an in vivo finite element model]. / Zahnimplantattestung am In-vivo-Finite-Elemente-Modell.

The finite element method (FEM) makes it possible to simulate biomechanical situations on a computer. In the present study the so-called voxel method [9, 14, 17, 18, 19] was used for the construction of the mandible model. For this, the relationship between the biological tissue (e.g. bone) and the corresponding attenuation coefficient of CT data (Hounsfield units = HU) were utilized. The CT data were obtained from an edentulous patient provided with a prosthesis borne on two titanium implants. In a parallel study, the bite forces of the same individual were measured. These were recorded digitally in three dimensions (cranio-caudal, anterior-posterior and left-right) The forces determined by a special program were then transferred to the FEM model implants. We were able to show that a bar joining the implants had a far greater effect on maximum equivalent stress than was expected from the measuring data alone. The highest stress at maximum occlusive force was lowered by 704% on using the connecting bar. On chewing, a stress reduction of 59.9% was observed. The reduction in stress achieved by the bar could, we believe, prolong the life of the implant.

Bone remodelling in humeral arthroplasty: follow-up using CT imaging and finite element modeling--an in vivo case study.

Little material is available in the literature about remodelling of the human humerus after implantation of a shoulder hemiarthroplasty. A 73-year-old patient was examined by CT 4 years after implantation of a right shoulder hemiarthroplasty, and the bone density as represented by Hounsfield values was compared with the contralateral side. Additionally, a three-dimensional finite-element model was generated from the image data and analysed. Bone density was reduced around the prosthesis when compared with the contralateral side. The stresses were transmitted through the prosthesis, while low bone stresses were found surrounding the prosthesis. Distally from the prosthesis, high stresses were found. On the control side, a more homogeneous stress distribution was noted. The results could be explained by bone resorption around the prosthesis caused by stress shielding; this hypothesis has to be confirmed by future studies.

Size and direction of hip joint forces associated with various positions of the acetabulum.

When total hip replacement is performed, the position of the acetabular component may affect wear and component survival time. We considered the questions: In what way does displacement of the hip joint center alter (1) the magnitude and (2) the direction of the resultant force? Biomechanical tests were carried out on a human multibody model. After displacement of the hip joint center, the resultant forces were calculated for the single leg stance. With the flexed single leg stance, maximum hip joint forces were observed with lateral, cranial, posterior displacement. The peak forces were affected by the modeling of a gluteus maximus wrapping point at the ischial tuberosity and were overestimated when this was removed. With the straight single leg stance, posterior displacement decreases the total load on the hip joint because of the increased leverage of the rectus femoris. With regard to the direction of the resultant force, medial displacement increases the angles in both planes, cranial displacement increases it in the sagittal plane (cranial, posterior-caudal, anterior), and anterior displacement decreases the angle in the sagittal plane and increases it in the frontal plane (medial, cranial-lateral, caudal). The direction of the force is relatively insensitive to displacement of the hip joint center. The results presented here indicate a marked increase in the force after lateral, cranial, posterior displacement of the center in the flexed single leg stance. To avoid extreme joint loading and to reduce the wear after total hip arthroplasty, the cranial and posterior regions of the acetabulum should be fully reconstructed. A high hip joint center has an adverse effect on the magnitude of the force, although the directions are hardly affected by it.

Mechanical effects of continuous passive motion on the lumbar spine in seating.

The aim of this study was to develop a model which describes the mechanical spinal response to small alternating pelvic stimulation induced by an active rotational movement of a normal chair. The rotary continuous passive motion (RCPM) of the seat about a vertical axis of only 0.6 degrees resulted in an increased in spinal length as opposed to the normal daily shrinkage, and back patients experienced pain relief. Passive and active exercies have been broadly applied for treating and healing spinal disorders. A rigid body package (ADAMS Android) was used to translate the stimulation of the ischial tubersoity in caudo-cranial handing-over visualisation. The parameters of the model were set so that the values of the global stiffness and geometry of the intervertebral discs could be changed. In vivo validation of the model was based on force and moment measurements using an internal AO fixator. The predicitons of the model concerning natural frequency (4.5Hz) in vertical direction and the axial torsion response on small pelvic torsion are comparable with experimental data.

Spinal load changes during rotatory dynamic sitting.

OBJECTIVE: To measure load and moment changes acting on the lumbar spine during rotatory sitting. BACKGROUND: A new chair concept generating dynamic stimuli by alternating rotations in the horizontal plane of the chair's seat was recently developed. METHODS: Load and moment changes were measured telemetrically with a spinal fixator device in vivo. RESULTS: A rotatory frequency of 0.22 Hz with an amplitude of 1.8 degrees to the right and left side showed maximum axial force changes in the fixator of 23 N and maximum bending moment changes of 0.52 Nm. CONCLUSIONS: Lumbar force and moment changes during dynamic sitting occur, although only one patient was included in the study. Reasons could be temporary muscular activation in order to adapt the body's equilibrium conditions at the end-point rotation. RelevanceOur measurements suggest that a rotatory chair does have an effect on lumbar spine forces. It becomes more likely that this concept could improve the discs' nutrition and may prevent low back pain.

Lumbar spine curvature during office chair sitting.

Prolonged sitting is generally accepted as a high risk factor in low back pain and it is frequently suggested that a lordotic posture of the lumbar spine should be maintained during sitting. We asked whether the sagittal curvature of the lumbar spine during sitting is affected by the seat tilt, backrest and the direction of the synchronised mechanism of the back and seat tilt (synchro tilt). Two office chairs were tested by multibody analysis interfacing a human model with a chair model. Results indicate that a synchronised mechanism of an office chair representing a posterior tilt of the seat while the backrest is reclined maintains an evenly distributed lumbar lordosis. The segmental angles are between 3.1 and 3.6 degrees at the lumbar vertebrae 1/2-4/5 (L1/2-L4/5). These lumbar spine segmental angles are not sensitive to the backrest height. In contrast, a synchro tilt concept with a reduction of the seat's posterior tilt while the backrest is reclined causes a strong reduction of the lumbar lordosis in backrest recline with a maximum reduction from 11.7 to 2.8 degrees in L4/5. As a consequence of these results, a synchro tilt concept with a posterior tilt of the seat while the backrest is reclined is preferable from the lumbar spine kinematics point of view.

[Stress analysis of an anatomically-adapted femur shaft prosthesis (Lubinus SPII)]. / Spannungsanalyse einer anatomisch adaptierten Femurschaftprothese (Lubinus SPII).

OBJECTIVE: Very good clinical long-term results of the Lubinus SP II hip prosthesis stem were reported in the literature. We therefore asked whether there is a relationship between these findings with biomechanical data of strain gauge measurements. METHOD: 14 strain gauges were applied at a femur being measured at 10 different load cases before and after implantation of the stem. RESULTS: After stem implantation a similar patterns of principal stress distributions was observed, however, the magnitude was markedly reduced. A striking reduction of the hoop stresses at the femoral calcar was seen in the case of a missing collar contact. Even in the case of a perfect collar contact the hoop stresses were diminished after strem implantation. The S-shaped physiological stem did not correspond with a specific stress pattern measured at the femoral surface. CONCLUSION: These results suggest that the stresses at the femoral calcar may be lower than the limits of bone growth while the other parts of the femur are more physiologically stressed. However, the prosthesis may tolerate a missing collar contact during a long follow-up period. The large experimental data file presented here could be used to validate future finite element analyses which could evaluate the stress distribution within internal parts of the bone and the cement layer.

Comparison of geometry-based and CT voxel-based finite element modelling and experimental validation.

The objectives of the paper presented here is the comparison of a geometry-based and voxel-based finite element (FE) method preprocessor of the human femur. The models were experimentally validated by strain gauge measurements (principal stress). The correlation coefficients (r) between the three methods (geometry-based FEM, voxel-based FEM, strain gauge measurements) were found to be in the range 0.91-0.94 (r2: 0.84-0.88). The relationships between the samples are highly significant (P = 0.001), where the strain gauge results are the independent variables. These results suggest that the validity with respect to the principal stress of a voxel-based modelling is similar to the validity of geometry-based modelling. In summary, therefore, we conclude that voxel-based meshing allows a straightforward interfacing with computerized tomography (CT) scans and might contribute to a clinically applicable FEM technology.

Centrifugal pumps as left ventricular assist for coronary revascularization on a beating heart.

During recent years, coronary bypass surgery has progressed toward minimizing invasiveness. One important feature of this approach is performing surgery on a beating heart. During the crucial phase of such surgery, the mechanical support of the heart with a left ventricular assist device (LVAD) is a possible option. During the period from October 1, 1994 until June 30, 1997, we employed a centrifugal pump system in 118 cases of coronary artery bypass graft (CABG) procedures with LVAD support (mechanically supported CABG [SUPPCAB]). A total of 179 distal anastomoses with an average of 1.5 +/- 0.5 coronary anastomoses per patient was performed. Three types of pumps were used: 23 BioPump, 87 Isoflow, and 8 Capiox systems. The median time on mechanical support was 44 min (range, 16-116 min). The mean flow rate during support time was 3.5 +/- 0.8 L/min, which results in a calculated flow of 1.7 +/- 0.6 L/min/m2 body surface area (BSA). The average flow was 3.2 +/- 0.8 L/min with the BioPump and 3.7 +/- 0.8 L/min with the Isoflow pump, respectively (p < 0.01). The mean arterial pressure during mechanical support was 75 +/- 12 mm Hg. In 2 patients, the pump system was kept running postoperatively in the ICU. Eight of the patients received operations under resuscitation or in cardiogenic shock. Nine (7.9%) of the patients did not survive the early postoperative phase. For coronary revascularization of the anterolateral and diaphragmatic parts of the heart, the SUPPCAB procedure is feasible with excellent mechanical support of the heart by centrifugal pumps. Especially in high risk cases, this procedure can be recommended.

Functions of hip joint muscles.

The relative torque components (percentages of total hip torque) of 26 hip muscles about three joint axes were calculated using a human multibody model to determine muscle actions as a function of hip joint angle. It was demonstrated that the muscle actions obtained from this straight line model of hip musculature were in good agreement with published functions; the data on muscle function changes during hip joint motions exceed the data available in the literature and could extend the anatomical knowledge which is available today.

[Fully automated generation of hip prosthesis voxel models of the femur for finite element analysis. Comparison of direct and subsequent prosthesis implantation]. / Die vollautomatische Erstellung implantatbestückter Voxelmodelle des Femurs zur FE-Analyse, Vergleich von direkter und nachträglicher Prothesenimplantation.

Two means of generating implant-bearing FE voxel models of the human femur are described and compared. The first method consists in the fully automatic production of a femur model from CT data and subsequent implantation of a prosthesis stem on the basis of a redefinition of existing voxel elements using geometrical data control. The second method processes the CT data of a femur already bearing a prosthesis. Here, the prosthesis elements are generated separately from the surrounding bony elements on the basis of the differing CT density thresholds. Although the two implantation algorithms produce comparable results, they have different fields of application. The method of numerical stress analysis involving automatic generation of voxel models can be applied to implant-bearing femurs. Clinical follow-up observations of patients with prostheses made of titanium can be supported by FE calculations.

Sensitivity of femoral strain pattern analyses to resultant and muscle forces at the hip joint.

An automated geometrical preprocessor was developed with the aim of creating three-dimensional finite element models (FEM) of the human femur. On the basis of postprocessed computed tomography data, this preprocessor makes possible rapid, flexible and regular meshing with 'brick' elements. Three different material properties were modelled at the present stage of development. Sensitivity analyses demonstrated that the strain energy density (SED) patterns of the different femoral parts were most sensitive to the implementation of an iliotibial tract force. The variation of the resultant hip force and abductor force direction within the sagittal plane demonstrated a SED minimum at an anterior inclination of 13 degrees; the variation of the resultant force direction within the frontal plane demonstrated a minimum SED at a medial inclination of 21 degrees relative to the mechanical axis of the lower limb. The orientation of the connecting line between the surface-SED-peaks in the horizontal view was found to be most sensitive to the variation of the resultant force within the sagittal plane.

[Use of voxel-oriented femur models for stress analysis. Generation, calculation and validation of CT-based FEM models]. / Die Anwendung voxelorientierter Femurmodelle zur Spannungsanalyse. Generierung, Berechnung und Validierung CT-basierter FEM-Modelle.

A means of interfacing computed tomography with the finite element method for the analysis of stress distribution in human femurs is described. With the aid of data transfer and image processing programs, three-dimensional voxel models of four macerated and a fresh-frozen cadaveric femur were generated from CT data sets. The underlying basic principle is direct generation of model elements from one or more adjacent CT-voxels, which are characterised by their coordinates and local absorption coefficients. A new algorithm was used for smoothing the edges of the model. All the models were analysed with respect to their stress distribution under different load situations, and the results were compared with the data of a parallel study utilizing the strain-gauge technique. Using the method of linear regression, a high correlation (coefficient of determination r2: 0.80-0.91) was found between the calculated and experimentally measured principal stresses at the surface of the bones.

[Predicting load bearing of the hip joint. Computerized analysis with a 3-D multibody model of the human]. / Belastungsvorhersagen am Hüftgelenk. Computeranalysen an einem 3-D-Mehrkörpermodell des Menschen.

Multibody analysis was applied to construct an advanced model of the human body, where the large joints and complete mass and inertial properties were implemented. The model represents the 50th-percentile rank of a male adult. The hip joint is controlled by three muscle forces. The muscle coordinates were taken from a data source, previously collected by our group. The model enables one to analyze 3D hip joint forces with respect to various joint angles and represents conceptually an improvement of the classical method of graphical statics, which was established by Pauwels [15]. A hip joint load of three times body weight was found in the single leg stance. A load of 3.7 times body weight was calculated when simulating a knee flexion angle of 90 degrees, and a ventral inclination of the resultant hip joint force was seen. A constant amount of gluteus medius muscle force was observed during flexion. An increasing flecting moment at the hip joint, however, had to be balanced by a significantly increased gluteus maximus muscle force. As a consequence, torsional forces can be studied by the system presented here and should also be considered when testing stems of hip prostheses. External muscle and joint forces are provided and can be used as input data for stress analyses.