Clinical classification of scoliosis patients using machine learning and markerless 3D surface trunk data.

Markerless 3D surface topography for scoliosis diagnosis and brace treatment can avoid repeated radiation known from standard X-ray analysis and possible side effects. Combined with the method of torso asymmetry analysis, curve severity and progression can be evaluated with high reliability. In the current study, a machine learning approach was utilised to classify scoliosis patients based on their trunk surface asymmetry pattern. Frontal X-ray and 3D scanning analysis with a clinical classification based on Cobb angle and spinal curve pattern were performed with 50 patients. Similar as in a previous study, each patient's trunk 3D reconstruction was used for an elastic registration of a reference surface mesh with fixed number of vertices. Subsequently, an asymmetry distance map between original and reflected torso was calculated. A fully connected neural network was then utilised to classify patients regarding their Cobb angle (mild, moderate, severe) and an Augmented Lehnert-Schroth (ALS) classification based on their full torso asymmetry distance map. The results reveal a classification success rate of 90% (SE: 80%, SP: 100%) regarding the curve severity (mild vs moderate-severe) and 50-72% regarding the ALS group. Identifying patient curve severity and treatment group was reasonably possible allowing for a decision support during diagnosis and treatment planning. Graphical abstract.

Innovative decision support for scoliosis brace therapy based on statistical modelling of markerless 3D trunk surface data.

Recently markerless 3D scanning methods receive an increased interest for therapy planning and brace treatment of patients with scoliosis. This avoids repeated radiation known from standard X-Ray analysis. Several authors introduced the method of asymmetry distance maps in order to classify curve severity and progression. The current work extends this approach by statistical mean shape 3D models of the human trunk in order to classify patients. 50 patients were included in this study performing frontal X-ray and 3D scanning analysis. All patients were classified by a clinician according to their Cobb angle and spinal curve pattern (Augmented-Lehnert-Schroth ALS). 3D reconstructions of each patient trunk were processed in a way to elastically register a reference surface mesh with fixed number of data points. Mean 3D shape models were generated for each curve pattern. An asymmetry distance map was then calculated for each patient and mean shape model. Single patient 3D reconstructions were classified according to severity and ALS treatment group. Optimal sensitivity and specificity was 97%/39% thoracic and 87%/42% lumbar respectively for detecting mild and moderate-severe patients. Identifying a treatment group was possible for three combined groups allowing to support decisions during diagnosis and therapy planning.

The influence of bone damage on press-fit mechanics.

Press-fitting is used to anchor uncemented implants in bone. It relies in part on friction resistance to relative motion at the implant-bone interface to allow bone ingrowth and long-term stability. Frictional shear capacity is related to the interference fit of the implant and the roughness of its surface. It was hypothesised here that a rough implant could generate trabecular bone damage during implantation, which would reduce its stability. A device was constructed to simulate implantation by displacement of angled platens with varying surface finishes (polished, beaded and flaked) onto the surface of an embedded trabecular bone cube, to different nominal interferences. Push-in (implantation) and Pull-out forces were measured and micro-CT scans were made before and after testing to assess permanent bone deformation. Depth of permanent trabecular bone deformation ('damage'), Pull-out force and Radial force all increased with implantation displacement and with implantation force, for all surface roughnesses. The proposed hypothesis was rejected, since primary stability did not decrease with trabecular bone damage. In fact, Pull-out force linearly increased with push-in force, independently of trabecular bone damage or implant surface. This similar behaviour for the different surfaces might be explained by the compaction of bone into the surfaces during push-in so that Pull-out resistance is governed by bone-on-bone, rather than implant surface-on-bone friction. The data suggest that maximum stability is achieved for the maximum implantation force possible (regardless of trabecular bone damage or surface roughness), but this must be limited to prevent periprosthetic cortical bone fracture, patient damage and component malpositioning.

Cartilage surface characterization by frictional dissipated energy during axially loaded knee flexion--an in vitro sheep model.

Cartilage defects and osteoarthritis (OA) have an increasing incidence in the aging population. A wide range of treatment options are available. The introduction of each new treatment requires controlled, evidence based, histological and biomechanical studies to identify potential benefits. Especially for the biomechanical testing there is a lack of established methods which combine a physiologic testing environment of complete joints with the possibility of body-weight simulation. The current in-vitro study presents a new method for the measurement of friction properties of cartilage on cartilage in its individual joint environment including the synovial fluid. Seven sheep knee joints were cyclically flexed and extended under constant axial load with intact joint capsule using a 6° of freedom robotic system. During the cyclic motion, the flexion angle and the respective torque were recorded and the dissipated energy was calculated. Different mechanically induced cartilage defect sizes (16 mm², 50 mm², 200 mm²) were examined and compared to the intact situation at varying levels of the axial load. The introduced setup could significantly distinguish between most of the defect sizes for all load levels above 200 N. For these higher load levels, a high reproducibility was achieved (coefficient of variation between 4% and 17%). The proposed method simulates a natural environment for the analysis of cartilage on cartilage friction properties and is able to differentiate between different cartilage defect sizes. Therefore, it is considered as an innovative method for the testing of new treatment options for cartilage defects.

Influence of cooling on curing temperature distribution during cementing of modular cobalt-chromium and monoblock polyethylene acetabular cups.

Total hip replacements for older patients are usually cemented to ensure high postoperative primary stability. Curing temperatures vary with implant material and cement thickness (30°C to 70°C), whereas limits for the initiation of thermal bone damage are reported at 45°C to 55°C. Thus, optimizing surgical treatment and the implant material are possible approaches to lower the temperature. The aim of this study was to investigate the influence of water cooling on the temperature magnitude at the acetabulum cement interface during curing of a modular cobalt-chromium cup and a monoblock polyethylene acetabular cup. The curing temperature was measured for SAWBONE and human acetabuli at the cement-bone interface using thermocouples. Peak temperature for the uncooled condition reached 70°C for both cup materials but was reduced to below 50°C in the cooled condition for the cobalt-chromium cup (P = .027). Cooling is an effective method to reduce curing temperature with metal implants, thereby avoiding the risk of thermal bone damage.

Dissipated energy as a method to characterize the cartilage damage in large animal joints: an in vitro testing model.

Several quantitative methods for the in vitro characterization of cartilage quality are available. However, only a few of these methods allow surgical cartilage manipulations and the subsequent analysis of the friction properties of complete joints. This study introduces an alternative approach to the characterization of the friction properties of entire joint surfaces using the dissipated energy during motion of the joint surfaces. Seven sheep wrist joints obtained post mortem were proximally and distally fixed to a material testing machine. With the exception of the carpometacarpal articulation surface, all joint articulations were fixed with 'Kirschner' wires. Three cartilage defects were simulated with a surgically introduced groove (16 mm(2), 32 mm(2), 300 mm(2)) and compared to intact cartilage without an artificial defect. The mean dissipated energy per cycle was calculated from the hysteresis curve during ten torsional motion cycles (±10°) under constant axial preload (100-900 N). A significant increase in dissipated energy was observed with increasing cartilage defect size and axial load (p<0.001). At lower load levels, the intact and 16 mm(2) defect showed a similar dissipated energy (p>0.073), while all other defect conditions were significantly different (p=0.015). All defect sizes were significantly different (p=0.049) at 900 N axial load. We conclude that the method introduced here could be an alternative for the study of cartilage damage, and further applications based on the principles of this method could be developed for the evaluation of different cartilage treatments.

Influence of peri-implant bone quality on implant stability.

INTRODUCTION: Insufficient primary stability is still reported for proximal humerus fractures in elderly patients. Fixation stability could be improved by aiming locking screws at bone volumes with better properties. The aims of this study were to investigate the bone regions engaged by the locking screws of a Proximal Humeral Nail (MultiLoc PHN), and to evaluate the influence of peri-screw bone quality on bone-nail construct stability. MATERIALS AND METHODS: Twelve cadaveric humeri were divided into two groups. The distal locking part of the PHN was fixed to the specimens. The nails were removed and the bones scanned using HR-pQCT. Bone properties were evaluated at the locations where the proximal locking screws would have been positioned after complete instrumentation. A three-part fracture model was used for mechanical testing of the instrumented bones, considering axial displacement and varus deformation as parameters of interest. RESULTS: The secondary locking screws targeted bone volumes in the posteromedial part of the humerus with statistically significant higher quality, thus reducing varus deformation. Significant correlation was found between axial displacement and bone properties at the primary proximal screws. Significant correlation was found between the varus deformation and apparent BMD at the secondary locking screws. CONCLUSION: The findings of this study confirmed that directing the proximal locking screws at bone regions with better properties can improve fixation stability.

Acromioclavicular joint reconstruction: a comparative biomechanical study of three techniques.

BACKGROUND: Acute acromioclavicular joint dislocations indicated for surgery can be treated with several stabilization techniques. This in vitro study evaluated the acromioclavicular joint stability after 3 types of validated repair techniques compared with the native situation. MATERIALS AND METHODS: Nine pairs (right-left) of intact cadaveric shoulder specimens were assigned to 3 study groups with randomly distributed samples according to the coracoclavicular distance. The groups were instrumented with acromioclavicular and coracoclavicular cerclages (CE), a Twin Tail TightRope (TR), or a locking compression superior and anterior clavicle plate (CP). Native and instrumented specimens were tested quasi-static nondestructively (superior: 70 N; anteroposterior: ± 35 N, 10 mm/min) and cyclically until failure (superior, valley load: 20 N; initial peak load: 70 N; increment: 0.02 N/cycle). RESULTS: The TR study group showed the highest (in N/mm) superoinferior (73.77 ± 14.04) and anteroposterior (29.58 ± 1.52) stiffness, followed by CE (superoinferior: 59.73 ± 10.33; anteroposterior: 24.31 ± 4.14) and CP (superoinferior: 24.08 ± 5.29). Instrumentation generally led to increased superoinferior and anteroposterior stiffness in each study group but to a significant superoinferior stiffness reduction for CP (P = .029). Significantly lower coracoclavicular displacement at valley load after 1 and 500 cycles was observed for TR (P = .018) and CE (P = .041) compared with CP. Cycles to failure were significantly higher in CE (7298 ± 1244 cycles, P = .011) and TR (4434 ± 727 cycles, P = .031) compared with CP (1683 ± 509 cycles). CONCLUSIONS: The CE and TR techniques led to similar biomechanical performances. The CE repair might mimic the native acromioclavicular joint stiffness better than the other 2 setups, leading to more physiological stabilization.

In vitro evaluation of the torsional strength reduction of neonate calf metatarsal bones with Bicortical defects resulting from the removal of external fixation implants.

OBJECTIVE: To compare the torsional strength of calf metatarsal bones with defects produced by removal of 2 different implants. STUDY DESIGN: In vitro mechanical comparison of paired bones with bicortical defects resulting from the implantation of 2 different external fixation systems: the transfixation pin (TP) and the pin sleeve system (PS). SAMPLE POPULATION: Neonatal calf metatarsal bones (n = 6 pairs). METHODS: From each pair, 1 bone was surgically instrumented with 2 PS implants and the contralateral bone with 2 TP implants. Implants were removed immediately leaving bicortical defects at identical locations between paired metatarsi. Each bone was tested in torque until failure. The mechanical variables statistically compared were the torsional stiffness, the torque and angle at failure, and work to failure. RESULTS: For TP and PS constructs, respectively, there were no significant differences between construct types for any of the variables tested. Mean ± SD torsional stiffness: 5.50 ± 2.68 and 5.35 ± 1.79 (Nm/°), P = .75; torque: 57.42 ± 14.84 and 53.43 ± 10.16 (Nm); P = .34; angle at failure: 14.76 ± 4.33 and 15.45 ± 4.84 (°), P = .69; and work to failure 7.45 ± 3.19 and 8.89 ± 3.79 (J), P = .17). CONCLUSIONS: Bicortical defects resulting from the removal of PS and TP implants equally affect the investigated mechanical properties of neonate calf metatarsal bones.

Comparative biomechanical evaluation of a pin-sleeve transfixation system in cadaveric calf metacarpal bones.

OBJECTIVE: To compare proximal fragment displacement and the peri-implant strain using a pin-sleeve cast (PSC) system and a transfixation pin cast (TPC) system on a cadaveric calf metacarpal bone fracture model. STUDY DESIGN: Experimental. SAMPLE POPULATION: Cadaveric calf metacarpal bones (n = 6 pairs). METHODS: Paired samples were instrumented with either the TPC or the PSC systems. Strain gauges were applied proximal to the transfixation implants and the bones encased in cast material. The distal part of the construct was removed to mimic an unstable distal comminuted fracture. Constructs were fixed to the material testing machine and initially loaded in axial compression in their elastic range to determine construct stiffness. Constructs were loaded cyclically with a sinusoidal curve that increased until failure. Variables compared statistically between constructs were the initial construct stiffness and, at given load points, the mean metacarpal axial displacement in loading and unloading condition and mean axial strain. RESULTS: Initial construct mean ± SD axial stiffness was not significantly different between constructs (PSC: 689 ± 258; TPC: 879 ± 306 N/mm). There was no significant difference between either investigated displacements of metacarpal bones transfixed with PSC and those transfixed with TPC at all load points. The PSC constructs had a significant decrease in the recorded mean strain (502 ± 340 µstrain) compared to the TPC construct (1738 ± 2218 µstrain). CONCLUSIONS: The PSC significantly reduced peri-implant strain with comparable axial displacement to the TPC in cadaveric calf metacarpal bones.

Biomechanical and computational evaluation of two loading transfer concepts for pancarpal arthrodesis in dogs.

OBJECTIVE: To evaluate 2 plate designs for pancarpal arthrodesis and their effects on load transfer to the respective bones as well as to develop a computational model with directed input from the biomechanical testing of the 2 constructs. SAMPLE: Both forelimbs from the cadaver of an adult castrated male Golden Retriever. PROCEDURES: CT imaging was performed on the forelimb pair. Each forelimb was subsequently instrumented with a hybrid dynamic compression plate or a castless pancarpal arthrodesis plate. Biomechanical testing was performed. The forelimbs were statically loaded in the elastic range and then cyclically loaded to failure. Finite element (FE) modeling was used to compare the 2 plate designs with respect to bone and implant stress distribution and magnitude when loaded. RESULTS: Cyclic loading to failure elicited failure patterns similar to those observed clinically. The mean ± SD error between computational and experimental strain was < 15% ± 13% at the maximum loads applied during static elastic loading. The highest bone stresses were at the distal extent of the metacarpal bones at the level of the screw holes with both plates; however, the compression plate resulted in slightly greater stresses than did the arthrodesis plate. Both models also revealed an increase in bone stress at the proximal screw position in the radius. The highest plate stress was identified at the level of the radiocarpal bone, and an increased screw stress (junction of screw head with shaft) was identified at both the most proximal and distal ends of the plates. CONCLUSIONS AND CLINICAL RELEVANCE: The FE model successfully approximated the biomechanical characteristics of an ex vivo pancarpal plate construct for comparison of the effects of application of different plate designs.

Biomechanical evaluation of two intramedullary nailing techniques with different locking options in a three-part fracture proximal humerus model.

BACKGROUND: Osteosynthesis of unstable proximal humerus fractures still remains challenging. The aim of this study was to investigate two intramedullary nailing techniques with different locking options in a three-part fracture model and prove whether two new fixation concepts, introducing additional locking screw-in-screws inserted through the head of the proximal screws, and a calcar screw, provide better stability. METHODS: A biomechanical testing model for three-part proximal humerus fractures including cyclic axial loading with increasing peak load and simultaneous pulling forces at the rotator cuff was used to test 12 pairs of human cadaver humeri, assigned to four groups and instrumented with either Targon PH (T1) or MultiLoc PHN in 3 different configurations (standard M1; two additional screw-in-screw M2; one additional calcar screw and two screw-in-screw M3). FINDINGS: Initial range of motion in internal-external rotation and mediolateral translation was smallest in M3 (1.82°; 0.11mm), biggest in T1 (3.63°; 0.51mm) and significantly different between these two groups (p=0.02 and p=0.04, respectively). M3 showed minimum head migration along the nail and varus tilting after 5000 cycles (0.31mm; 0.20°) and 10000 cycles (1.59mm; 0.34°). M2 and M3 performed better than M1 and T1 regarding varus collapse. The highest number of cycles to failure was observed for M3 (20733) and the lowest for T1 (10083) with significant difference between these two groups (p=0.04). INTERPRETATION: The configuration with two screw-in-screw and a calcar screw was superior in most aspects. The screw-in-screws were found to contribute against varus collapse. Both new fixation concepts could provide better stability in proximal humerus fractures.

Bone marrow modified acrylic bone cement for augmentation of osteoporotic cancellous bone.

The use of polymethylmethacrylate (PMMA) cement to reinforce fragile or broken vertebral bodies (vertebroplasty) leads to extensive bone stiffening. This might be one reason for fractures at the adjacent vertebrae following this procedure. PMMA with a reduced Young's modulus may be more suitable. The goal of this study was to produce and characterize PMMA bone cements with a reduced Young's modulus by adding bone marrow. Bone cements were produced by combining PMMA with various volume fractions of freshly harvested bone marrow from sheep. Porosity, Young's modulus, yield strength, polymerization temperature, setting time and cement viscosity of different cement modifications were investigated. The samples generated comprised pores with diameters in the range of 30-250 µm leading to porosity up to 51%. Compared to the control cement, Young's modulus and yield strength decreased from 1830 to 740 MPa and from 58 to 23 MPa respectively by adding 7.5 ml bone marrow to 23 ml premixed cement. The polymerization temperature decreased from 61 to 38 ∘C for cement modification with 7.5 ml of bone marrow. Setting times of the modified cements were lower in comparison to the regular cement (28 min). Setting times increased with higher amounts of added bone marrow from around 16-25 min. The initial viscosities of the modified cements were higher in comparison to the control cement leading to a lower risk of extravasation. The hardening times followed the same trend as the setting times. In conclusion, blending bone marrow with acrylic bone cement seems to be a promising method to increase the compliance of PMMA cement for use in cancellous bone augmentation in osteoporotic patients due to its modified mechanical properties, lower polymerization temperature and elevated initial viscosity.

Influence of interface condition and implant design on bone remodelling and failure risk for the resurfaced femoral head.

Resurfacing of the femur has experienced a revival, particularly in younger and more active patients. The implant is generally cemented onto the reamed trabecular bone and theoretical remodelling for this configuration, as well as uncemented variations, has been studied with relation to component positioning for the most common designs. The purpose of this study was to investigate the influence of different interface conditions, for alternative interior implant geometries, on bone strains in comparison to the native femur, and its consequent remodelling. A cylindrical interior geometry, two conical geometries and a spherical cortex-preserving design were compared with a standard implant (ASR, DePuy International, Ltd., UK), which has a 3° cone. Cemented as well as uncemented line to line and press-fit conditions were modelled for each geometry. A patient-specific finite element model of the proximal femur was used with simulated walking loads. Strain energy density was compared between the reference and resurfaced femur, and input into a remodelling algorithm to predict density changes post-operatively. The common cemented designs (cylindrical, slightly conical) had strain shielding in the superior femoral head (>35% reduction) as well as strain concentrations (strain>5%) in the neck regions near the implant rim. The cortex-preserving (spherical) and strongly conical designs showed less strain shielding. In contrast to the cemented implants, line to line implants showed a density decrease at the centre of the femoral head, while all press-fit versions showed a density increase (>100%) relative to the native femur, which suggests that uncemented press-fit implants could limit bone resorption.

Does impaction matter in hip resurfacing? A cadaveric study.

Eight pairs of fresh frozen human femora were prepared for hip resurfacing. One side of each pair was impacted gently, the other side vigorously. After implantation procedure, specimens were loaded in a material testing machine to the ultimate fracture load. Median impaction loads on the vigorously implanted side were 11,298N compared to 1374N on the gently implanted side. Failure loads in the high-impact group (median, 8873N) were significantly (P = .0078) reduced when compared with the low-impact group (median, 9237N). The study stresses that meticulous reaming of the femoral head and the pinhole is of tremendous importance. Remaining obstacles can lead to excessive loads, while attempting to enforce the correct seating of the implant. Only careful, slight tapping should be applied to ensure final seating.

Primary stability of uncemented femoral resurfacing implants for varying interface parameters and material formulations during walking and stair climbing.

Primary stability of uncemented resurfacing prosthesis is provided by an interference fit between the undersized implant and the reamed bone. Dependent on the magnitude of interference, the implantation process causes high shear forces and large strains which can exceed the elastic limit of cancellous bone. Plastification of the bone causes reduced stiffness and could lead to bone damage and implant loosening. The purpose in this study was to determine press-fit conditions which allow implantation without excessive plastic bone deformation and sufficient primary stability to achieve bone ingrowth. In particular, the influence of interference, bone quality and friction on the micromotion during walking and stair-climbing was investigated. Therefore elastic and plastic finite element (FE) models of the proximal femur were developed. Implantation was realized by displacing the prosthesis onto the femur while monitoring the contact pressure, plastic bone deformation as well as implantation forces. Subsequently a physiologic gait and stair-climbing cycle was simulated calculating the micromotion at the bone-implant interface. Results indicate that plastic deformation starts at an interference of 30microm and the amount of plastified bone at the interface increases up to 90% at 150microm interference. This effect did not reduce the contact pressure if interference was below 80microm. The micromotion during walking was similar for the elastic and plastic FE models. A stable situation allowing bony ingrowth was achieved for both constitutive laws (elastic, plastic) for walking and stair climbing with at least 60microm press-fit, which is feasible with clinically used implantation forces of 4kN.