Douglas D. Robertson

Three-dimensional analysis of the proximal part of the humerus: relevance to arthroplasty.

Background: Recreation of normal anatomical relationships may be important to optimize the outcome of proximal humeral arthroplasty. With use of computerized tomographic data and three-dimensional computer modeling, we concurrently studied both extramedullary and intramedullary humeral morphology, including canal shape, and related these findings to the design of proximal humeral prostheses.Methods: Sixty cadaveric humeri (thirty pairs: fifteen from male donors and fifteen from female donors) were studied. Three-dimensional computer models were built from canal and periosteal contours extracted from computerized tomographic data and multiple measured anatomical parameters, including humeral canal axis, humeral head center, and hinge point offset; greater tuberosity and bicipital groove offset; humeral head center, radius, thickness, retroversion, and inclination; and size and torsion of sections of the canal.Results: On the average, the humeral head center was offset both medially (seven millimeters) and posteriorly (two millimeters) from the humeral axis. The humeral head hinge point did not line up with the axis but instead was laterally offset by an average of seven millimeters. The average humeral head thickness was nineteen millimeters. The humeral head thickness and length were proportionately linked. There was marked variability in all of these parameters. Humeral head inclination averaged 41 degrees but was less variable than previously described, with 95 percent of our sample within the range of 35 to 46 degrees. The proximal section of the humeral canal was retroverted, and the retroversion was found to be similar to that of the humeral head on statistical analysis. Version of the middle and distal sections of the canal, however, was dissimilar to that of the proximal section of the canal. Proximal humeral retroversion was found to be extremely variable and averaged 19 degrees. The accuracy, reliability, and repeatability of the computer-based-model measurements were found to be excellent.Conclusions: Measurements of external proximal humeral morphology made with three-dimensional computer models of cadaveric specimens derived from the Midwestern United States agreed, in general, with those described for different populations evaluated with different measuring techniques. Proximal humeral morphology was extremely variable as highlighted by the large ranges of measurements seen for all variables. Examination of the intramedullary morphology showed that there is an internal version, with measurements dependent on the canal distance distal to the anatomical neck.Clinical Relevance: Because of the marked variabilities seen in proximal humeral morphology, newer prosthetic designs are now allowing surgeons to control multiple prosthetic variables. An understanding of the normal values for proximal humeral morphology can serve as an important guideline for component selection, especially when the normal anatomy is distorted. Additionally, variations in intramedullary version may have important consequences for future designs of press-fit proximal humeral replacement.

Total Hip Prosthesis Metal-artifact Suppression Using Iterative Deblurring Reconstruction

PURPOSE CT of total joint prostheses is limited by metal artifact produced mainly by missing projection data. Iterative deblurring reconstruction is less sensitive to missing projection data than filtered backprojection. A software CT simulator was used to compare total hip prosthesis images reconstructed using standard filtered backprojection, filtered backprojection after linear interpolation of missing data, and iterative deblurring. METHOD Unilateral and bilateral total hip replacements with metal-backed or all-polyethylene acetabular prostheses were simulated using bone, metal, and polyethylene annuli and circles of metal and water (soft tissue). Material attenuation properties were taken from the literature. The simulation assumed that no X-rays penetrated the metal. Simulated projection data were reconstructed using filtered backprojection, filtered backprojection after linear interpolation of missing data, and iterative deblurring. Visual observations and objective region-of-interest analyses were made. RESULTS Even with no X-rays penetrating the metal, iterative deblurring produced almost no visible artifact within the bone or soft tissues. Bone edge detection and sizing were more easily and accurately done from the iterative deblurring images. All reconstruction techniques underestimated bone and water CT attenuation values. Metal artifact was worse for the bilateral or metal-backed prostheses. CONCLUSION Iterative deblurring generated nearly metal-artifact-free images in this simulation. Filtered backprojection, even after linear interpolation, produced typical clinical metal-artifact images.

Evaluation of CT techniques for reducing artifacts in the presence of metallic orthopedic implants

Metallic intramedullary orthopedic implants generate artifacts that can markedly degrade transaxial CT images. The artifacts, typically seen as starburst streaking, result primarily from reconstructions involving missing projection data. Two approaches are clinically available to reduce the artifacts around orthopedic implants. These are (a) the imaging of implants with lower attenuation coefficients or smaller path lengths (less attenuating objects) and (b) the planar reformatting of image data. The sizing accuracy of these two approaches was quantified using phantoms and the efficacy using cadaveric femoral specimens. Results demonstrated that metal artifacts may be reduced and accurate bony dimensional data obtained.

A mathematical model for the evaluation of the behaviour during flexion of condylar-type knee prostheses

A 3D knee model was developed in order to evaluate the mechanical behaviour during flexion of condylar-type knee prosthesis. Based on the total energy minimization principle, it takes into account the articular surfaces (the tibial surface being deformable), the body weight, and the patello femoral joint. It generates the kinematics of the joint, the motion of the centre of contact, the quadriceps forces, the pressure distribution on the tibial plateau, and ligament lengths and forces between 0 and 120 degrees of flexion. The results for ten digitized knees and the commercially available prostheses are presented. They are in general agreement with experimental results published in the literature. It is concluded that this computer program may be, within its limitations, a useful tool in the preliminary evaluation of new condylar-type knee prosthesis designs.

Bone loss in the distal anterior femur after total knee arthroplasty

Bone loss in the distal anterior femur in asymptomatic total knee arthroplasty (TKA) patients has been noted roentgenographically and during revision surgery. A retrospective roentgenographic review of 147 TKA cases was carried out to document bone loss. The influence that the mode of fixation (porous coated and cemented) and the implant design have on bone loss was examined. The time of onset and the progression of bone loss were studied. Bone loss occurred in the distal anterior femur in the majority of cases reviewed (68%). The prevalence of bone loss was independent of the mode of fixation and the implant design. By qualitative observation, roentgenographically detectable bone loss occurred within the first postoperative year and did not progress further. Previously three-dimensional finite element analysis demonstrated that the replacement of the bearing surface of the femur with a stiff metallic implant reduces the stress in the distal anterior femur by at least one order of magnitude. It is therefore speculated that the observed bone loss results from stress shielding. The apparent lack of progression may reflect the development of a new remodeling equilibrium under the altered stress conditions. The bone loss in the distal anterior femur described has not been implicated as a source of failure. However, since the bone strength in the femoral region is compromised as it becomes osteopenic, bone failure may occur with longer periods of cyclic loading. Furthermore, as a result of bone loss, revision arthroplasty may be more difficult.

Design and fabrication of cementless hip stems.

Theoretical and experimental studies before and after stem insertion demonstrated that stresses and strains were closer to normal for uncemented stems than for cemented stems. The values were affected by relative tightness of fit in the proximal and distal regions and by the presence or absence of a collar. For designing an optimal fit-stem, the average femoral geometry was first determined. The stem was used in a photoelastic coating study that showed a continuous strain field over the bone surface. The average proximomedial bone strains for intact, press-fit, loose press-fit with collar, and proximally cemented (to simulate ingrowth) designs were 100%, 65%, 101%, and 54%, respectively. Localized patches of high strain were seen on the proximal bone surface and beneath a collar, indicative of localized stem-bone contact points.

Femoral deformity in adults with developmental hip dysplasia.

Quantitative computed tomography and 3 dimensional modeling were used to portray the deformity of the proximal femur in 24 Japanese adults with low subluxations to high dislocations secondary to developmental dysplasia of the hip. Periosteal and canal bony contours were extracted, 3 dimensional models generated, and morphologic parameters were calculated for each femur. Three dimensional illustrations of the average deformity and variability were created. Morphologic parameters were not found to be statistically correlated with the degree of the disease. Interestingly, the major axis of the canal contours of the proximal femur was found to be aligned with the plane of the femoral neck (anteversion), regardless of the degree of anteversion. Thus, the amount of version correctable in an uncemented prosthesis is limited, and at times may require a special prosthesis, overreaming, undersizing and cementing, or an osteotomy. Additionally, the proximal medial curvature of the dysplastic femurs was straighter than that of normal femurs. This necessitated a corresponding reduction in the proximal medial curvature of a conventional uncemented prosthesis to match the medial curvature of the individual femur and the average developmentally dysplastic femur. This objective description of the developmentally dysplastic femur corroborates clinical observations, highlights some unrecognized findings, provides a rationale for planning reconstructions, and aids in the design of prostheses for adult patients with this deformity.

Design of Custom Hip Stem Prostheses Using Three-dimensional Ct Modeling

Long life expectancy, demand for high activity levels, and bone loss at the time of revision motivate the search for reliable and successful nonce-mented hip stem designs. It is hypothesized that improved implant fit may increase the longevity of noncemented total joints. Quantitative X-ray CT has enabled the use of a computerized stem design program, which designs an optimal-fit hip stem for individual femurs. Computed tomography and interactive image processing methods are used to generate the individual three-dimensional femoral models, which are used by the stem design program. Optimal-fit design provides maximum stem-bone contact while satisfying the requirement of being surgically insertable. Previous methods of custom implant design, including those that use three-dimensional CT modeling, have not provided optimal stem-bone fit. Quantitative results of this new process are presented.

Distal loss of femoral bone following total knee arthroplasty. Measurement with visual and computer-processing of roentgenograms and dual-energy x-ray absorptiometry.

The changes in bone-mineral content in the distal aspect of a cadaveric femur that had been prepared for insertion of the femoral component of a total condylar knee prosthesis were evaluated with visual inspection and computer-processing of roentgenograms and with dual-energy x-ray absorptiometry. Seventeen small slices of bone, each three millimeters wide, were removed so that, finally, 89 per cent of the bone was lost from the distal femoral metaphysis. Standardized lateral roentgenograms of the specimen were made with use of a reference step-wedge of hydroxyapatite, and dual-energy x-ray absorptiometry studies were performed with the x-ray beam tangential to the interface. The roentgenograms were digitized and the bone mineral was measured with use of computer analysis. Dual-energy x-ray absorptiometry was performed with and without the femoral prosthesis in place, in order to determine the effect of the metallic prosthesis on the accuracy of the measurement. A bone loss of 25 per cent or more was identified visually by all five of the readers 100 per cent of the time; losses of 20 to 24 per cent, 15 to 19 per cent, 10 to 14 per cent, and 3 to 9 per cent were correctly identified 92, 75, 66, and 59 per cent of the time, respectively. The measurements of bone-mineral content that were obtained from the digitized roentgenograms were linearly correlated with the actual bone-mineral content (the ash content) (r = 0.97, p < 0.001) and were three times more accurate than the visual readings. The determinations of bone-mineral content with dual-energy x-ray absorptiometry correlated highly with the ash content (r = 1.00, p < 0.001) and were seven times more accurate than the visual readings. There was only a 4 per cent difference between the measurements with dual-energy x-ray absorptiometry made with the prosthesis in place and those made without it in place (p < 0.01). Dual-energy x-ray absorptiometry was the most accurate of the three methods and could detect the smallest experimentally created loss; computer-processing and visual-processing of roentgenograms detected losses of 8 per cent or more and 25 per cent or more, respectively. Dual-energy x-ray absorptiometry and computer-processing of the roentgenograms quantified the bone loss, while visual-processing could detect only the presence or absence of bone loss.(ABSTRACT TRUNCATED AT 400 WORDS)

Strategies for improving fixation of femoral components in total hip arthroplasty.

As strategies are considered for improving fixation of femoral components in total hip arthroplasty (THA), one is challenged to exceed the standard set by contemporary cement procedures. However, despite the improved ten- to 15-year clinical results anticipated with current cementing techniques, the limited fatigue strength of polymethylmethacrylate warrants continued investigation of alternative systems, particularly for younger patients and in revision arthroplasty. Design considerations for femoral stems for cementless THA include (1) initial mechanical stability afforded by the stem shape, (2) strength and stiffness of the stem, and (3) surface features relating to biocompatibility and attachment to bone. In one approach a fit-and-fill algorithm has been implemented to design stems that maximize contact between prosthesis and cortex in priority areas to achieve stability. Titanium is recommended for the fabrication of such stems because of its corrosion resistance, its biocompatibility, and its modulus, which is lower than that of cobalt-chromium alloy. Long-term fixation of these implants will be dependent upon the maintenance of normal strain patterns in the host bone. Achievement of this goal will require additional strategies that combine optimal fit and optimal material properties of the prosthesis.