Intraoperative planning of knee replacement components, targeting a desired functional outcome, requires a calibrated patient-specific model of the patient's soft-tissue anatomy and mechanics. Previously, a surgical technique was demonstrated for measuring knee joint kinematics and kinetics consistent with modern navigation systems in conjunction with the development of a patient-customizable knee model. A data efficient approach for the model calibration task was achieved utilizing the sensitivity of the model to simulated clinical hand manipulations of the knee joint requiring 85% less computations. For this numerical investigation a simplified knee joint model, based on the OpenKnee repository, consisting of bone (rigid), cruciate ligaments (single-bundle, nonlinear spring), collateral ligaments (multiple nonlinear springs), articular cartilage (rigid, pressure-over-closure relationship), and combined capsule/meniscus (linear springs) was created using a custom Matlab (MathWorks)-Abaqus (Dassault Systèmes) implicit finite element modeling framework (Figure 1). A sensitivity analysis was performed by applying constant loading along the anterior-posterior, medial-lateral, varus-valgus, and internal-external directions (30 N for forces and 3 Nm for moments) while perturbing each customizable parameter positively and negatively by 1 mm at 0, 25, 50, 75 and 100 degrees of flexion. A constant load of 150 N was maintained in compression. The change in static endpoint position was measured relative to the respective position without perturbation. Sensitivity results were then arranged by load direction and principal component analysis was subsequently performed (Table 1). First a single optimization task was simulated including all model parameters and all loading sequences with the goal of minimizing the kinematic differences between the reference model and a perturbed model (Figure 2). Second, a piecewise optimization task was designed using only the sensitive parameters for a spanning set of loads for the same perturbed model. Parameters 3 and 4 were tuned using internal and external endpoints. Then parameters 1 and 5 were tuned using the anterior endpoints. Similarly, parameters 2 and 7 were tuned using the posterior endpoints. Finally, parameter 8 was tuned using the varus endpoints. All loadings were observed to be insensitive to parameter 6 (ACL-Y). The number of model evaluations required were 2520 and 390 for the single and piecewise optimizations, respectively. The single simulation task recovered all parameters within 0.57 mm on average compared to 0.64 mm on average for the piecewise task. Kinematic errors due to the calibration technique were within 0.001 mm and 0.18 deg compared to 0.001 mm and 0.04 deg. Computational cost for the optimization task required to calibrate a patient-specific knee model was reduced while maintaining clinically relevant accuracy. This model reduction approach will further enable the rapid adoption of the technology for intraoperative planning of knee replacement components based on targeted functional outcomes.
Tibial and femoral component malalignment is poorly tolerated in uni- and bi-compartmental knee replacement. Poor outcomes may still occur while using navigation or robotic-assisted bone preparation, which currently require surgeon assessment to establish a preoperative plan for implant placement. Choosing where to place partial knee replacement components is a challenging task that depends on complicated interactions between patient variability and implant design. We developed a patient-customizable knee model that can assist surgeons by providing a quantitative measure of knee laxity. In order to build upon previous knee modeling efforts and to demonstrate the technique, three-dimensional femur and tibia bone and articular cartilage geometries were obtained from the OpenKnee finite element repository ( The model was run through a series of simulated passive flexion paths. At each degree of flexion, combinations of anterior-posterior and medial-lateral forces as well as internal-external and varus/valgus moments were applied and the resulting joint kinematics were recorded. These results represent the passive envelope of knee motion, which is used to characterize knee laxity. An optimization framework was developed to iteratively tune the cruciate ligament model to match a virtual set of passive loading conditions. A majority of preoperative planning techniques only monitor geometric targets such as flexion and extension gaps, limb alignment, restoration of the joint line, and tibial component slope. Patient-customized knee models can be tuned to quantify post-operative knee laxity and identify the range of tolerable alignment of partial knee replacement components. Future work will employ in-vitro testing to validate the capability of the model to identify patient-specific cruciate ligament parameters.
There is strong current interest to provide reliable treatments for one- and two-compartment arthritis in the cruciate-ligament intact knee. An alternative to total knee arthroplasty is to resurface only the diseased compartments with discrete compartmental components. Placing multiple small implants into the knee presents a greater surgical challenge than total knee arthroplasty, and it is not certain natural knee mechanics can be maintained. The goal of this study was to compare functional kinematics in cruciate-intact knees with either medial unicondylar (mUKA), mUKA plus patellofemoral (mUKA+PF), or bi-unicondylar (biUNI) arthroplasty using discrete compartmental implants with preparation and placement assisted by haptic robotic technology. Nineteen patients with 21 knee arthroplasties consented to participate in an I.R.B. approved study of knee kinematics with a cruciate-retaining multicompartmental knee arthroplasty system. All subjects presented with knee OA, intact cruciate ligaments, and coronal deformity ranging from 7° varus to 4° valgus. All subjects received multicompartmental knee arthroplasty using haptic robotic-assisted bone preparation an average of 13 months (6–29 months) before the study. Eleven subjects received mUKA, five subjects received mUKA+PF, and five subjects received biUKA. Subjects averaged 62 years of age and had an average body mass index of 31. Combined Knee Society Pain/Function scores averaged 102 ± 28 preoperatively and 169 ± 26 at the time of study. Knee range of motion averaged −3° to 120° preoperatively and −1° to 129° at the time of the study. Knee motions were recorded using video-fluoroscopy while subjects performed step-up/down, kneeling and lunging activities. The three-dimensional position and orientation of the implant components were determined using model-image registration techniques (Fig. 1). The AP locations of the medial and lateral condyles were determined by computing a distance map between the femoral condyles and the tibial articular surfaces.INTRODUCTION
METHODS
There is great interest in technologies to improve the accuracy and precision in placing implants for total hip arthroplasty (THA). Malik et al. (J Arthroplasty, 2010) showed that an imageless navigation system could be used to produce accurate measures of acetabular cup alignment compared to a CT-based alignment method using an imaging phantom. In this study we sought to compare the precision of an image-based navigation system with post-operative CT scans in a clinical patient cohort who received navigation-assisted THA. Eighteen patients with 20 hips consented to this IRB-approved analysis of intra- and post-operative THA cup alignment. All patients received THA with image-assisted alignment (MAKO Surgical, Fort Lauderdale). Nominal cup placement, subject to intraoperative surgeon adjustment and approval, was 40° radiographic inclination (RI) and 20° radiographic anteversion (RA) according to the definitions of Murray (JBJS-Br, 1993). Intraoperative cup alignment was measured by collecting five points on the cup rim with an optically tracked stylus. Postoperative cup alignment was measured by registering pre- and post-operative pelvic models generated from CT scans, and determining the postoperative cup orientation relative to the pre-operative pelvis coordinate system (Figure 1). Repeated measures testing of the CT-based measurements on 10 patient scans showed precision and bias of 0.7° and 0° for radiographic inclination, and 0.6° and 0.1° for radiographic anteversion.Introduction
Methods
Adjusting joint gaps and establishing mediolateral (ML) soft tissue balance are considered essential interventions for better outcomes in total knee arthroplasty (TKA). However, the relationship between intraoperative laxity measurements and weight-bearing knee kinematics has not been well explored. The goal of this study was to establish how intraoperative joint gaps and ML soft tissue balance affect postoperative kinematics in posterior-stabilized (PS)-TKA. We investigated 44 knees with 34 patients who underwent primary PS-TKA. Subjects averaged 71 ± 7 years at the time of surgery, included 8 male and 36 female knees with a preoperative diagnosis of osteoarthritis in 38 knees and rheumatoid arthritis in 6 knees. A single surgeon performed all the surgeries with mini-midvastus approach. After independent bone cutting, soft tissues were released on a case-by-case basis to obtain ML balance. The femoral trial and a tensor were put in place, and the patella was reduced to the original position. A joint distraction force of 40 lb was applied by the tensor, and the central joint gaps and ML tilting angles were measured at 0°, 10°, 30°, 60°, 90°, 120° and 135° flexion (Fig. 1). We defined a “gap difference” as a gap size difference between one gap and another, which represents the gap change between the two knee flexion positions. ML soft tissue balance was assessed by measuring the mean joint gap tilting angle over all flexion angles for each patient. Based on the tilting angle, the 44 knees were classified into three groups: The knees with the mean joint gap tilting of less than −1.0° (13 knees), between −1.0 and 1.0° (14 knees), and over 1.0° (17 knees). At least 1.5 year after surgery, a series of dynamic squat radiographs and 3 static lateral radiographs of straight-leg standing, lunge at maximum flexion, and kneeling at maximum flexion, were taken for each patient. The 3-dimensional position and orientation of the implant components were determined using model-based shape matching techniques (Fig. 2). Correlations between intraoperative measurements and knee kinematics were analyzed. The knee kinematics was also compared among three tilting groups.Introduction:
Methods:
There is great contemporary interest to provide treatments for knees with medial or medial plus patellofemoral arthritis that allow retention of the cruciate ligaments and the natural lateral compartment. Options for bicompartmental arthroplasty include custom implants, discrete compartmental implants and monoblock off-the-shelf implants. Each approach has potential benefits. The monoblock approach has the potential to provide a cost-efficient off-the-shelf solution with relatively simple surgical instrumentation and procedure. The purpose of this study was to determine if monoblock bicompartmental knee arthroplasty shows evidence of retained cruciate ligament function and clinical performance more similar to unicompartmental arthroplasty than total knee arthroplasty. Nine females and one male patient were enrolled in this IRB approved study. Each subject received unilateral bicompartmental knee arthroplasty an average of 2.6 years (2.0 to 3.6 years) prior to this study. Subjects averaged 65 years (58–72 years) and 28 BMI (25–31) at the time of surgery. Mean outcome scores at the time of study were 97/95 for the Knee Society knee/function score, 16.4 Oxford score, 6.5 UCLA Activity score and 137 degrees range of motion. Subjects were observed using dynamic fluoroscopy during lunge, kneeling and step-up/down activities. Subjects also received CT scans of the knee in order to create bone/implant composite shape models. Model-image registration techniques were used to determine 3D knee kinematics (Figure 1). Knee angles were quantified using a flexion-abduction-rotation Cardan sequence and condylar translations were determined from the lowest point on the condyle with respect to the transverse plane of the tibial segment. Maximum knee flexion during lunge and kneeling activities averaged 112°±8° and 125°±7°, respectively. Tibial internal rotation averaged 10°±6° and 12°±10° for the lunge and kneeling activities. For both deeply flexed postures, the medial condyle was 1 mm anterior to the AP center of the tibia while the lateral condyle was 11 mm and 13 mm posterior to the tibial center. For the step-up/down activity, tibial internal rotation increased an average of 2° from 5° to 75° flexion, but was quite variable (Figure 2). Medial condylar translations averaged 4 mm posterior from 5° to 25° flexion, followed by 6 mm anterior translation from 25° to 80° flexion (Figure 3). All knees showed posterior condylar translation from extension to early flexion. An important potential benefit to any bicompartmental arthroplasty treatments is retention of the cruciate ligaments and maintenance of more natural knee function. The knees in this study showed excellent or good clinical outcomes and functional scores, and relatively activity high levels. There was no evidence of so-called paradoxical anterior femoral translation during early flexion, indicating retained integrity of the natural AP stabilizing structures. Weight-bearing deep flexion during lunge and kneeling activities was comparable to previously reported unicompartmental and well-performing total knee arthroplasty subjects. Kinematics were quite variable between subjects. Monoblock bicompartmental arthroplasty appears to permit functional retention of the cruciate ligaments, consistent with functionally stable knees. Further efforts should focus on the specific surgical placement of off-the-shelf bicompartmental implants to optimize knee function and provide consistent knee mechanics.
There is great interest to provide repeatable and durable treatments for arthritis localized to one or two compartments in the cruciate-ligament intact knee. We report a series of efforts to develop and characterize an implant system for partial knee resurfacing. We studied distal femoral morphology and found that the sagittal-plane relationships between the condylar and trochlear surfaces are highly variable (Figs 1 and 2). In response, we report the design of a multi-compartmental system of implants intended to anatomically resurface any combination of compartments (Fig 3). Finally, we report the results of a pilot fluoroscopic study of the in vivo knee kinematics in patients who received medial, medial plus patellofemoral and bi-condylar knee arthroplasty. The kinematic results suggest these treatments provide a stable knee with intact cruciate ligament function. This work shows various partial knee resurfacing treatments have the potential to provide excellent knee mechanics and clinical outcomes.
Total hip arthroplasty (THA) is regarded as one of the most successful surgeries in medicine. However, recent studies have revealed that ideal acetabular cup implantation is achieved less frequently than previously thought, as little as 50% of the time. It is well known that malalignment of the acetabular component in THA may result in dislocation, reduced range of motion, or accelerated wear. This study reports accuracy of a tactile robotic arm system to ream the acetabulum and impact an acetabulur cup compared to manual instrumentation. 12 fresh frozen cadaveric acetabulae were pre-operatively CT scanned and 3D templating was used to plan the center of rotation, and anteversion and inclination of the cup. Each specimen received THA, six prepared manually and six prepared with robotic arm guidance. Tactile, visual, and auditory feedback was provided through robotic guidance as well as navigated guided reaming and cup impaction. The robotic guidance constrained orientation of instruments thus constraining anteversion, inclination, and center of rotation for reaming, trialing, and final cup impaction. Post-operative CT scans were taken of each specimen to determine final cup placement for comparison to the pre-operative plans.INTRODUCTION
METHODS
Unicompartmental knee arthroplasty (UKA) can achieve excellent clinical and functional results for patients suffering from single compartment osteoarthritis. However, UKA is considered to be more technically challenging to perform, and malalignment of the implant components has been shown to significantly contribute to UKA failures. The purpose of this investigation was to determine the clinically realized accuracy of UKA component placement using surgical navigation and dynamically referenced tactile-robotics. Pre-op CT, post-op CT, and surgical plan were available for 22 knees out of the first 45 procedures performed using a new tactile-guided robotic system. 3D component placement accuracy was assessed by comparing the pre-operative plan with the post-operative implant placement (desired versus actual). Bone and implant models were obtained from postoperative CT scans taken immediately following the surgery. A 3D to 3D iterative closest point registration procedure was performed and the measured implant position was directly compared to the preoperative plan. Errors were assessed as single axis root-mean-square (RMS) entities.INTRODUCTION
METHODS
We introduce the concept of total knee arthroplasty (TKA) fingerprinting as a tool to characterize and graphically convey the sensitivity of a TKA design to surgical variability in implant component position and patient-related anatomic factors. Identifying sensitive directions preoperatively which would cause undesirable effects may decrease revision surgery by informing surgical decisions and planning. To provide several examples of TKA fingerprinting, we estimated and compared the contact forces in a single TKA type for several configurations, simulating surgical variability and patient-related anatomical factors during a loaded deep squat. The purpose of this study is not to analyze the behavior of this specific TKA design but rather to illustrate a tool that could be used to show, in general, how surgical errors or anatomical factors can alter patello-femoral (PF) and tibio-femoral (TF) contact forces compared to its own reference configuration. Computed tomography images of one full cadaveric leg were used to generate 3D models of the bones and to obtain a physiological knee model assuming standard positions of the main soft tissue insertions. A fixed bearing posterior stabilized knee TKA design was considered in this study. The prosthesis was a medium size, replaced both cruciate ligaments and resurfaced the patella. Following standard surgical procedure, the TKA was virtually implanted, thus defining its reference configuration. Each derivative replaced knee model was then obtained by changing the values of one parameter, or a combination of two, in a range based on literature and surgical experience (Table 1). A 10 s loaded squat to 120° was performed for each configuration, with a constant vertical hip load of 200 N. These settings match the experimental tests performed in a previous in-vitro analysis on cadaver legs. Each replaced model was developed and analyzed using a validated musculoskeletal modeling software. The model of the knee included TF contacts and PF contacts of the TKA components, passive soft tissues and active muscle elements. The external forces (ground reaction and weights), the muscle forces (quadriceps and hamstrings) and the frictional forces are applied to the knee joint through the machine. The mechanical properties of the tissues were obtained from literature. With these settings, for each model, both the maximum PF and TF contact forces have been evaluated.Introduction
Materials and methods
Unicompartmental knee arthroplasty (UKA) is an increasingly attractive and clinically successful treatment for individuals with isolated medial compartment disease who demand high levels of function. A major challenge with UKA is to place the components accurately so they are mechanically harmonious with the retained joint surfaces, ligaments and capsule. Misalignment of UKA components compromises clinical outcomes and implant longevity. Cobb et al. (JBJS-Br 2006) showed that robot-assisted placement of UKA components was more accurate than traditional techniques, and subsequently that the clinical outcomes were improved. Cobb’s method, however, employed rigid intraoperative stabilization of the bones in a stereotactic frame, which is impractical for routine clinical use. Robotic systems have now advanced to include dynamic bone tracking technologies so that rigid fixation is no longer required. The question is -Do these robotic systems with dynamic bone tracking provide the same accuracy advantages demonstrated with robotic systems with rigidly fixed bones? We compared robot-assisted and traditionally instrumented UKA in six bilateral pairs of cadaver specimens. In all knees, a CT-based preoperative plan was performed to determine the ideal positions and orientations for the implant components. Traditional manual instruments were utilized with a tissue-sparing approach to implant one knee of each pair. A haptic robotic system acting as a virtual cutting guide was used to perform the robot-assisted UKA, again with a tissue-sparing approach. Postoperative CT scans were obtained from all knees, and the 3D placement errors were quantified using 3D-to-3D registration of implant and bone models to the reconstructed CT volumes. The magnitudes of femoral implant orientation error were significantly smaller for the robot-assisted implants compared to traditionally implanted components (4° vs 11°, p<
0.001), but the magnitudes of femoral placement error did not reach significance (3mm vs. 5mm, p=0.056). The magnitudes of tibial implant placement error were not significantly different (4mm vs. 5mm and 7° vs. 7°, p>
0.05). Well-placed UKA implants can provide durable and excellent functional results, which is an increasingly attractive option for young and active patients with severe compartmental osteoarthritis who wish not to have or to delay a total knee replacement. Previous studies have demonstrated significant improvement in implant placement accuracy and clinical results with robot-assisted surgery using rigid bone fixation. This study demonstrates it is possible to achieve significant accuracy improvements with robot-assisted techniques allowing free bone movement. Additional larger trials will be required to determine if these differences are realized in clinical populations.
