When a knee flex deeply, the posterior side of thigh and calf contact. The contact force is unignorable to estimate the load acting on a knee because the force generates extensional moment on the knee, and the moment might be about 20–80% of the flexional moment generated by a floor reacting force. Besides, the thigh-calf contact force varies so much even if the posture or the test subject are the same that it is hard to use the average value to estimate the knee load. We have assumed that the force might change not only by the individual physical size but also by a slight change of the posture, especially the angle of the upper body. Therefore we tried to create the estimation equation for the thigh-calf contact force using both anthropometric sizes and posture angles as parameters. The objective posture was kneeling, both plantarflexing and dorsiflexing the ankle joint. Test subjects were 10 healthy males. They were asked to sit on a floor with kneeling, and to tilt their upper body forward and backward. The estimation equations were created as the linear combinations of the parameters, determining the coefficient as to minimize the root mean square errors. We used the parameters as explanatory variables which could be divided into posture parameters and individual parameters. Posture parameters included the angle of upper body, thigh and lower thigh. Individual parameters included height, weight, axial and circumferential lengths of thigh and lower thigh. The magnitude of the force was normalized by a body weight, and the acting position was expressed by the moment arm length around a knee joint and normalized by a height. As a result, the adjusted coefficient of determination improved and the root mean square error decreased when using both posture and individual parameters, though there were large errors when neglecting either parameters. The accuracy decreased little when using the same equation for plantarflexed and dorsiflexed kneeling in magnitude. The relation of measured and estimated values of the magnitude and acting position, using the common equation with all the parameters. It might be because the difference of the postures could be described by the inclination angle of a thigh. In both postures, the magnitude of a thigh-calf contact force was mainly affected by the posture and acting position by the individual parameters. When calculating the knee joint load, the errors would be about 8.59 Nm on the knee moment and 290 N on the knee load when using just an average, and they would decrease to 2.23 Nm and 74 N respectively using the estimation equation

The spinopelvic alignment is often assessed via the Pelvic Incidence-Lumbar Lordosis (PI-LL) mismatch. Here we describe and validate a simplified method to evaluating the spinopelvic alignment through the L1-Pelvis angle (L1P). This method is set to reduce the operator error and make the on-film measurement more practicable. 126 standing lateral radiographs of patients presenting for Total Hip Arthroplasty were examined. Three operators were recruited to label 6 landmarks. One operator repeated the landmark selection for intra-operator analysis. We compare PI-LL mismatch obtained via the conventional method, and our simplified method where we estimate this mismatch using PI-LL = L1P - 90°. We also assess the method's reliability and repeatability. We found no significant difference (p > 0.05) between the PI-LL mismatch from the conventional method (mean 0.22° ± 13.6) compared to L1P method (mean 0.0° ± 13.1). The overall average normalised root mean square error (NRMSE) for PI-LL mismatch across all operators is 7.53% (mean -3.3° ± 6.0) and 6.5% (mean -2.9° ± 4.9) for the conventional and L1P method, respectively. In relation to intra-operator repeatability, the correlation coefficients are 0.87 for PI, 0.94 for LL, and 0.96 for L1P. NRMSE between the two measurement sets are PI: 9.96%, LL: 5.97%, and L1P: 4.41%. A similar trend is observed in the absolute error between the two sets of measurements. Results indicate an equivalence in PI-LL measurement between the methods. Reproducibility of the measurements and reliability between operators were improved. Using the L1P angle, the classification of the sagittal spinal deformity found in the literature translates to: normal L1P<100°, mild 100°<L1P<110°, and severe L1P>110°. Surgeons adopting our method should expect a small improvement in reliability and repeatability of their measurements, and a significant improvement of the assessment of the mismatch through the visualisation of the angle L1P

Introduction & Aims. In other medical fields, smart implantable devices are enabling decentralised monitoring of patients and early detection of disease. Despite research-focused smart orthopaedic implants dating back to the 1980s, such implants have not been adopted into regular clinical practice. The hardware footprint and commercial cost of components for sensing, powering, processing, and communicating are too large for mass-market use. However, a low-cost, minimal-modification solution that could detect loosening and infection would have considerable benefits for both patients and healthcare providers. This proof-of-concept study aimed to determine if loosening/infection data could be monitored with only two components inside an implant: a single-element sensor and simple communication element. Methods. The sensor and coil were embedded onto a representative cemented total knee replacement. The implant was then cemented onto synthetic bone using polymethylmethacrylate (PMMA). Wireless measurements for loosening and infection were then made across different thicknesses of porcine tissue to characterise the sensor's accuracy for a range of implantation depths. Loosening was simulated by taking measurements before and after compromising the implant-cement interface, with fluid influx simulated with phosphate-buffered saline solution. Elevated temperature was used as a proxy for infection, with the sensor calibrated wirelessly through 5 mm of porcine tissue across a temperature range of 26–40°C. Results. Measurements for loosening and infection could be acquired simultaneously with a duration of 4 s per measurement. For loosening, the debonded implant-cement interface was detectable up to 10 mm with 95% confidence. For temperature, the sensor was calibrated with a root mean square error of 0.19°C at 5 mm implantation depth and prediction intervals of ±0.38°C for new measurements with 95% confidence. Conclusions. This study has demonstrated that with only two onboard electrical components, it is possible to wirelessly measure cement debonding and elevated temperature on a smart implant. With further development, this minimal hardware/cost approach could enable mass-market smart arthroplasty implants

INTRODUCTION. To assess and compare the effect of new orthopedic surgical procedures, in vitro evaluation remains critical during the pre-clinical validation. Focusing on reconstruction surgery, the ability to restore normal kinematics and stability is thereby of primary importance. Therefore, several simulators have been developed to study the kinematics and create controlled boundary conditions. To simultaneously capture the kinematics in six degrees of freedom as outlined by Grood & Suntay, markers are often rigidly connected to the moving bone segments. The position of these markers can subsequently be tracked while their position relative to the bones is determined using computed tomography (CT) of the test specimen with the markers attached. Although this method serves as golden standard, it clearly lacks real-time feedback. Therefore, this paper presents the validation of a newly developed real-time framework to assess knee kinematics at the time of testing. MATERIALS & METHODS. A total of five cadaveric fresh frozen lower limb specimens have been used to quantitatively assess the difference between the golden standard, CT based, method and the newly developed real-time method. A schematic of the data flow for both methods. Prior to testing, both methods require a CT scan of the full lower limb. During the tests, the proximal femur and distal tibia are necessarily resected to fit the knees in the test setup, thus also removing the anatomical landmarks needed to evaluate their mechanical axis. Subsequently, a set of three passive markers are rigidly attached to the femur and tibia, referred to as M3F and M3T respectively. For the CT based method, the marker positions are captured during the tests and a second CT scan is eventually performed to link the marker positions to the knee anatomy. Using in-house developed software, this allowed to offline evaluate the knee kinematics in six degrees of freedom by combining both CT datasets with the tracked marker positions. For the newly developed real-time method, a calibration procedure is first performed. This calibration aims to link the position of the 3D reconstructed bone and landmarks with the attached markers. A set of bone surface points is therefore registered. These surface points are obtained by tracking the position of a pen while touching the bone surface. The pen's position is thereby tracked by three rigidly attached markers, denoted M3P. The position of the pen tip is subsequently calculated from the known pen geometry. The iterative closest point (ICP) algorithm is then used to match the 3D reconstructed bone to the registered surface points. Two types of 3D reconstructions have therefore been considered. First, the original reconstructions were used, obtained from the CT data. Second, a modified reconstruction was used. This modification accounted for the finite radius (r = 1.0 mm) of the registration pen, by shifting the surface nodes 1.0 mm along the direction of the outer surface normal. During the tests, the positions of the femur and tibia markers are tracked and streamed in real-time to an in-house developed, Matlab based software framework (MathWorks Inc., Natick, Massachussets, USA). This software framework simultaneously calculates the bone positions and knee kinematics in six degrees of freedom, displaying this information to the surgeons and operators. To assess the accuracy, all knee specimens have been subjected to passive flexion-extension movement ranging from 0 to 120 degrees of flexion. For each degree of freedom, the average root mean square (RMS) difference between both measurement methods has been evaluated during this movement. In addition, the distribution of the registered surface points has been assessed along the principal directions of the uniformly meshed 3D reconstructions (average mesh size of 1.0 mm). RESULTS. The root mean square difference between both measurements indicates a strong dependency on the variance of the registered points. This dependency is particularly pronounced when using the original 3D reconstructions in combination with the ICP algorithm, with an R. 2. = 0.76 and 0.85 for the translational and rotational degrees of freedom respectively. When using the modified 3D reconstructions, which compensates for the finite radius of the marker tip, this dependency becomes negligible (R. 2. = 0.10 and 0.05). Using this modified 3D reconstruction, the average difference between both measurements is also reduced to an average value of 1.20 degrees and 1.47 mm. DISCUSSION. The difference in kinematic parameters between both measurement techniques is an order of magnitude lower than the claimed accuracy of the motion tracking cameras. However, the difference is in line with the inter- and intra- observer variability when identifying bony landmarks around the knee. Since these landmarks are essential to calculate knee kinematics, it is understood that the proposed real-time system is sufficiently accurate to study these kinematics

INTRODUCTION. Statistical shape models (SSM) have become a common tool to create reference models for design input and verification of total joint implants. In a recent discussion paper around Artificial Intelligence and Machine Learning, the FDA emphasizes the importance of independent test data [1]. A leave-one-out test is a standard way to evaluate the generalization ability of an SSM [2]; however, this test does not fulfill the independence requirement of the FDA. In this study, we constructed an SSM of the knee (femur and tibia). Next to the standard leave-one-out validation, we used an independent test set of patients from a different geographical region than the patients used to build the SSM. We assessed the ability of the SSM to predict the shapes of knees in this independent test set. METHODS. A dataset of 82 computed tomography (CT) scans of Caucasian patients (42 male, 40 female) from 11 different geographic locations in France, Germany, Austria, Italy and Australia were used as training set to make an SSM of the femur and tibia. A leave-one-out test was performed to assess the ability of the SSM to predict shapes within the training set. A test dataset of 4 CT scans of Caucasian patients from Russia were used for the validation. The SSM was fitted onto each of the femur and tibia shapes and the root mean square error (RMSE) was measured. RESULTS. The leave-one-out tests showed that the femur and tibia SSMs were able to predict patients in the input population with an RMSE of 0.59 ± 0.1 mm (average ± standard deviation) for the femur and 0.70 ± 0.1 mm for the tibia. The validation test showed that the femur and tibia SSMs were able to predict the shapes of the Russian patients with an RMSE 0.62 ± 0.1 mm for the femur and 0.71 ± 0.1 mm for the tibia. DISCUSSION. There were no significant differences in the ability of the SSM to predict femur and tibia shapes of patients in a new geographic region compared to the ability of the SSM to predict shapes within the training set. CONCLUSIONS. Based on this study, 11 different geographic locations in France, Germany, Austria, Italy and Australia provide a complete sample of the Caucasian population. Using an independent set of CT scans is a valuable tool to further validate the generalization ability of an SSM. For any figures or tables, please contact authors directly

INTRODUCTION. While computational models have been used for many years to contribute to pre-clinical, design phase iterations of total knee replacement implants, the analysis time required has limited the real-time use as required for other applications, such as in patient-specific surgical alignment in the operating room. In this environment, the impact of variation in ligament balance and implant alignment on estimated joint mechanics must be available instantaneously. As neural networks (NN) have shown the ability to appropriately represent dynamic systems, the objective of this preliminary study was to evaluate deep learning to represent the joint level kinetic and kinematic results from a validated finite element lower limb model with varied surgical alignment. METHODS. External hip and ankle boundary conditions were created for a previously-developed finite element lower limb model [1] for step down (SD), deep knee bend (DKB) and gait to best reproduce in-vivo loading conditions as measured on patients with the Innex knee (. orthoload.com. ) (Figure1). These boundary conditions were subsequently used as inputs for the model with a current fixed-bearing total knee replacement to estimate implant-specific kinetics and kinematics during activities of daily living. Implant alignments were varied, including variation of the hip-knee-ankle angle-±3°, the frontal plane joint line −7° to +5°, internal-external femoral rotation ±3°, and the tibial posterior slope 5° and 0°. Through varying these parameters a total of 2464 simulations were completed. A NN was created utilizing the NN toolbox in MATLAB. Sequence data inputs were produced from the alignment and the external boundary conditions for each activity cycle. Sequence outputs for the model were the 6 degree of freedom kinetics and kinematics, totaling 12 outputs. All data was normalized across the entire data set. Ten percent of the simulation runs were removed at random from the training set to be used for validation, leaving 2220 simulations for training and 244 for validation. A nine-layer bi-long short-term memory (LSTM) NN was created to take advantage of bi-LSTM layers ability to learn from past and future data. Training on the network was undertaken using an RMSprop solver until the root mean square error (RMSE) stopped reducing. Evaluation of NN quality was determined by the RMSE of the validation set. RESULTS. The trained NN was able to effectively estimate the validation data. Average RMSE over the kinetics of the validation data set was 140.7N/N∗m while the average RMSE over the kinematics of the validation data set was 4.47mm/deg (Figure 2,3–DKB, Gait shown). It is noted the error may be skewed by the larger magnitude kinetics and kinematics in the DKB activity as the average RMSE for just SD and gait was 85.9N/N∗m and 2.8mm/deg for the kinetics and kinematics, respectively. DISCUSSION. The accuracy of the generated NN indicates its potential for use in real-time modeling, and further work will explore additional changes in post-operative soft-tissue balance as well as scaling to patient-specific geometry

Introduction. Machine learning is a relatively novel method to orthopaedics which can be used to evaluate complex associations and patterns in outcomes and healthcare data. The purpose of this study is to utilize 3 different supervised machine learning algorithms to evaluate outcomes from a multi-center international database of a single shoulder prosthesis to evaluate the accuracy of each model to predict post-operative outcomes of both aTSA and rTSA. Methods. Data from a multi-center international database consisting of 6485 patients who received primary total shoulder arthroplasty using a single shoulder prosthesis (Equinoxe, Exactech, Inc) were analyzed from 19,796 patient visits in this study. Specifically, demographic, comorbidity, implant type and implant size, surgical technique, pre-operative PROMs and ROM measures, post-operative PROMs and ROM measures, pre-operative and post-operative radiographic data, and also adverse event and complication data were obtained for 2367 primary aTSA patients from 8042 visits at an average follow-up of 22 months and 4118 primary rTSA from 11,754 visits at an average follow-up of 16 months were analyzed to create a predictive model using 3 different supervised machine learning techniques: 1) linear regression, 2) random forest, and 3) XGBoost. Each of these 3 different machine learning techniques evaluated the pre-operative parameters and created a predictive model which targeted the post-operative composite score, which was a 100 point score consisting of 50% post-operative composite outcome score (calculated from 33.3% ASES + 33.3% UCLA + 33.3% Constant) and 50% post-operative composite ROM score (calculated from S curves weighted by 70% active forward flexion + 15% internal rotation score + 15% active external rotation). 3 additional predictive models were created to control for the time required for patient improvement after surgery, to do this, each primary aTSA and primary rTSA cohort was subdivided to only include patient data follow-up visits >20 months after surgery, this yielded 1317 primary aTSA patients from 2962 visits at an average follow-up of 50 months and 1593 primary rTSA from 3144 visits at an average follow-up of 42 months. Each of these 6 predictive models were trained using a random selection of 80% of each cohort, then each model predicted the outcomes of the remaining 20% of the data based upon the demographic, comorbidity, implant type and implant size, surgical technique, pre-operative PROMs and ROM measures inputs of each 20% cohort. The error of all 6 predictive models was calculated from the root mean square error (RMSE) between the actual and predicted post-op composite score. The accuracy of each model was determined by subtracting the percent difference of each RMSE value from the average composite score associated with each cohort. Results. For all patient visits, the XGBoost decision tree algorithm was the most accurate model for both aTSA & rTSA patients, with an accuracy of ∼89.5% for both aTSA and rTSA. However for patients with 20+ month visits only, the random forest decision tree algorithm was the most accurate model for both aTSA & rTSA patients, with an accuracy of ∼89.5% for both aTSA and rTSA. The linear regression model was the least accurate predictive model for each of the cohorts analyzed. However, it should be noted that all 3 machine learning models provided accuracy of ∼85% or better and a RMSE <12. (Table 1) Figures 1 and 2 depict the typical spread and RMSE of the actual vs. predicted total composite score associated with the 3 models for aTSA (Figure 1) and rTSA (Figure 2). Discussion. The results of this study demonstrate that multiple different machine learning algorithms can be utilized to create models that predict outcomes with higher accuracy for both aTSA and rTSA, for numerous timepoints after surgery. Future research should test this model on different datasets and using different machine learning methods in order to reduce over- and under-fitting model errors. For any figures or tables, please contact the authors directly

Osteophytes are bony spurs on normal bone that develop as an adaptive reparative process due to excessive stress at/near a joint. As osteophytes develop from normal bone, they are not always well depicted in common imaging techniques (e.g. CT, MRI). This creates a challenge for preoperative planning and image-guided surgical methods that are commonly incorporated in the clinical routine of orthopaedic surgery. The study examined the accuracy of osteophyte detection in clinical CT and MRI scans of varying types of joints. The investigation was performed on fresh-frozen ex-vivo human resected joints identified as having a high potential for presentation of osteophytes. The specimens underwent varying imaging protocols for CT scanning and clinical protocols for MRI. After dissection of the joint, the specimens were subjected to structured 3D light scanning to establish a reference model of the anatomy. Scans from the imaging protocols were segmented and their 3D models were co-registered to the light scanner models. The quality of the osteophyte images were evaluated by determining the Root Mean Square (RMS) error between the segmented osteophyte models and the light scan model. The mean RMS errors for CT and MRI scanning were 1.169mm and 1.419mm, respectively. Comparing the different CT parameters, significance was achieved with scanning at 120kVp and 1.25mm slice thickness to depict osteophytes; significance was also apparent at a lower voltage (100kVp). Preliminary results demonstrate that osteophyte detection may be dependent on the degree of calcification of the osteophyte. They also illustrate that while some imaging parameters were more favourable than others, a more accurate osteophyte depiction may result from the combination of both MRI and CT scanning

Introduction. Partial knee arthroplasty (PKA) has demonstrated the potential to improve patient satisfaction over total knee arthroplasty. It is however perceived as a more challenging procedure that requires precise adaptation to the complex mechanics of the knee. A recently developed PKA system aims to address these challenges by anatomical, compartment specific shapes and fine-tuned mechanical instrumentation. We investigated how closely this PKA system replicates the balance and kinematics of the intact knee. Materials and Methods. Eight post-mortem human knee specimens (age: 55±11 years, BMI: 23±5, 4 male, 4 female) underwent full leg CT scanning and comprehensive robotic (KUKA KR140 comp) assessments of tibiofemoral and patellofemoral kinematics. Specimens were tested in the intact state and after fixed bearing medial PKA. Implantations were performed by two experienced surgeons. Assessments included laxity testing (anterior-posterior: ±100 N, medial-lateral: ±100 N, internal-external: ±3 Nm, varus- valgus: ±12 Nm) under 2 compressive loads (44 N, 500 N) at 7 flexion angles and simulations of level walking, lunge and stair descent based on in-vivo loading profiles. Kinematics were tracked robotically and optically (OptiTrack) and represented by the femoral flexion facet center (FFC) motions. Similarity between intact and operated curves was expressed by the root mean square of deviations (RMSD) along the curves. Group data were summarized by average and standard deviation and compared using the paired Student's T-test (α = 0.05). Results. During the varus-valgus balancing assessment the medial and lateral opening of the PKAs closely resembled the intact openings across the full arch of flexion, with RMSD values of 1.0±0.5 mm and 0.4±0.2 mm respectively. The medial opening was nearly constant across flexion, its average was not statistically different between intact (3.8±1.0 mm) and PKA (4.0±1.1 mm) (p=0.49). Antero-posterior envelope of motion assessments revealed a close match between the intact and PKA group for both compression levels. Net rollback was not statistically different, either under low compression (intact: 10.9±1.5 mm, PKA: 10.7±1.2, p=0.64) or under high compression (intact: 13.2±2.3 mm, PKA: 13.0±1.6 mm, p=0.77). Similarly, average laxity was not statistically different, either under low (intact: 7.7±3.2 mm, PKA: 8.6±2.5 mm, p=0.09) or under high (intact: 7.2±2.6 mm, PKA: 7.8±2.2 mm, p=0.08) compression. Activities of daily living exhibited a close match in the anterior-posterior motion profile of the medial condyle (RMSD: lunge: 2.2±1.0 mm, level walking: 2.4±0.9 mm, stair descent: 2.2±0.6 mm) and lateral condyle (RMSD: lunge: 2.4±1.4 mm, level walking: 2.2±1.4 mm, stair descent: 2.7±2.0 mm). Patellar medial-lateral tilt (RMSD: 3.4±3.8°) and medial-lateral shift (RMDS: 1.5±0.6 mm) during knee flexion matched closely between groups. Conclusion. Throughout the comprehensive functional assessments the investigated PKA system behaved nearly identical to the intact knee. The small residuals are unlikely to have a clinical effect; further studies are necessary as cadaveric studies are not necessarily indicative of clinical results. We conclude that PKA with anatomical, compartment specific shapes and fine-tuned mechanical instrumentation can be adapted precisely to the complex mechanics of the knee and replicates intact knee balance and kinematics very closely

Objective. This paper aims to analyze the kinetics of the over-ground wheel-type body weight supporting system (BWS); tendency changes of low extremity joint moment (hip, knee, ankle), 3 axis accelerations of a trunk, cadence and gait velocity as weight bearing level changes. Method. 15 subjects (11 males, 4 females, age:23.63.5, height:170.65.1cm, weight:69.0210.75kg) who had no history of surgery participated. 6 levels (0%, 10%, 20%, 30%, 40% and 50%) of BWS were given to subjects at self-selected gait velocity and kinetic data was calculated using a motion capture system, Vicon. ®. (Vicon, UK). Results. Maximum joint moments at the hip, knee, and ankle decrease as weight bearing increases on the sagittal plane. However, no significant decrease was found after 20% level of BWS at the hip and knee joint. On the other hand, the maximum ankle joint moment keeps decreasing. The root mean square (RMS) values of the acceleration in three directions: anterior-posterior (AP), medial-lateral (ML), and vertical(V) are analyzed. All 3-dimensional accelerations decrease as BWS increases while there is no significant difference over 20% level of BWS in the ML acceleration. V acceleration is reduced almost by half as soon as BWS level starts, but no further significant decrease can be found after 30% level of BWS. The AP acceleration tends to keep decreasing as BWS level increases. The cadence and gait velocity with wheel-type BWS decreases as BWS increases. Discussion. The maximum joint moments of the hip and knee do not significantly decrease when BWS exceeds a certain level, which is different from the case with BWS on treadmill; the maximum moments tend to keep decreasing linearly as BWS level increases on treadmill. In the case of the hip joint, the maximum moment is generated between toe-off and pre-swing phase, which generates force to push a trunk forward. With higher BWS, forward progression of the trunk is assisted by the wheel rather than driven by the lower extremity. It should be noticed that not only the tendency is different from BWS on treadmill, but the magnitude of the maximum hip moment is smaller than that of BWS on treadmill when BWS level is over 20%. The maximum knee joint moment is generated at the loading-response phase working as braking and shock absorption during gait, and thus the decrease in the maximum knee moment implies that less braking and shock absorption are required as BWS level increases. Only the maximum ankle joint torque keeps decreasing as BWS increases. The ankle moment is considered the largest contributor to forward acceleration. The tendency of the maximum ankle moment and the AP acceleration are similar (to what?) as weight bearing proceeds, which implies that walking speed slows down with the wheel-type BWS; the cadence is also reduced as BWS increases. Conclusion. The results highlight the difference of wheel-type BWS from BWS on treadmill, and provide information on how BWS level affects the joint moment and gait patterns. These outcomes can be utilized as a guideline of gait rehabilitation for people with lower-limb musculoskeletal impairments

In computer assisted orthopaedic surgery, intraoperative registration is commonly performed by fitting features acquired from the exposed bone surface to a preoperative virtual model of the bone geometry. In cases where the acquired spatial measurements are unreliable or have been inappropriately chosen, the registration result can degenerate. Current performance indicators, such as the root mean squared (RMS) error and the spatial distribution of the registered feature errors may not be sufficient to warn the surgeon of such a case. In this study, statistical analysis is applied to the registration outcomes of perturbed variants of a collected point set. In this way, it is possible to assess the ability of the original set to represent the underlying surface, taking into account the distribution of the points as well as errors introduced during the acquisition process. Confidence measures are calculated to predict the reliability of the original registration result and therefore the robustness of the point set itself. For proof of concept, this method has been tested in simulation with a CT-generated tibia model. The algorithm was used to identify the 10 best performing of a population of 1000 randomly generated point sets. All registration outcomes produced by these point sets were found to be superior to those resulting from sets of the same size produced manually using an optimised point-acquisition protocol. Preliminary results suggest that this method, alongside the standard RMS and residual point error distribution, may be used to provide the surgeon with a reliable indication of registration outcome in the operating room

In professional football a key factor regarding injury is the time to return to play. Accurate prediction of this would aid planning by the club in the event of injury. It would also aid the club medical staff. Gaussian processes may be used for machine learning tasks such as regression and classification. This study determines whether machine-learning methods may be used for predicting how many days a player is unavailable to play. A database of injuries at one English Premier League Professional Football Club was reviewed for a number of factors for each injury. Twenty-five variables were recorded for each injury, including time to return to play. This was determined to be the response variable. We used a Gaussian process model with a Laplacian kernel to determine whether the return to play could be predicted from the other variables. The root mean square error was 13.186 days (S.D.: 8.073), the mean absolute error was 8.192 days (S.D.:13.106) and the mean relative error 171.97% (S.D.:75.56%). A linear trend was observed and the model demonstrated high accuracy with greater errors being observed for cases where the value of the response variable was higher, i.e. in those cases where the time to return to play was lengthy. This is the first step in attempting to design a computer-based model that will accurately predict the time for a professional footballer to return to play. The model is extremely accurate for most cases, with errors increasing as the severity of the case increases too

Introduction. Joint mechanics and implant performance have been shown to be sensitive to ligament properties [1]. Computational models have helped establish this understanding, where optimization is typically used to estimate ligament properties for recreation of physically measured specimen-specific kinematics [2]. If available, contact metrics from physical tests could be used to improve the robustness and validity of these predictions. Understanding specimen-specific relationships between joint kinematics, contact metrics, and ligament properties could further highlight factors affecting implant survivorship and patient satisfaction. Instrumented knee implants offer a means to measure joint contact data both in-vivo and intra-operatively, and can also be used in a controlled experimental environment. This study extends on previous work presented at ISTA [3], and the purpose here was to evaluate the use of instrumented implant contact metrics during optimization of ligament properties for two specimens. The overarching goal of this work is to inform clinical joint balancing techniques and identify factors that are critical to implant performance. Methods. Total knee arthroplasties were performed on 4 (two specimens modeled) cadeveric specimens by an experienced orthopaedic surgeon. An instrumented trial implant (VERASENSE, OrthoSensor, Inc., Dania Beach, FL) was used in place of a standard insert. Experimentation was performed using a simVITROTM controlled robotic musculoskeletal simulator (Cleveland Clinic, Cleveland, OH) to apply intra-operative style loading and measure tibiofemoral kinematics. Three successive laxity style tests were performed at 10° knee flexion: anterior-posterior force (±100 N), varus-valgus moment (±5 Nm), and internal-external moment (±3 Nm). Tibiofemoral kinematics and instrumented implant contact metrics were measured throughout testing (Fig. 1). Specimen-specific finite element models were developed for two of the tested specimens and solved using Abaqus/Explicit (Dassault Systèmes). Relevant ligaments and rigid bone geometries were defined using specimen-specific MRIs. Virtual implantation was achieved using registration and each ligament was modeled as a set of nonlinear elastic springs (Fig. 1). Stiffness values were adopted from the literature [2] while the ligament slack lengths served as control variables during optimization. The objective was to minimize the root mean square difference between VERASENSE measured tibiofemoral contact metrics and the corresponding model results (Fig. 1). Results and Discussion. The models for both specimens successfully recreated joint kinematics with average errors less than 4° in rotations, and 3 mm in translations (not shown). Minus a systematic offset in θ for specimen 3, AFD and θ contact kinematics also realized good agreement for both specimens (Fig. 2). Contact forces were generally over-predicted, though both specimens recreated the experimental trends (Fig. 2). The present work shows continued progress towards simulation based tools that can be used for both research and to support the clinical decision making process. A separate ISTA submission presents assessment of these model's predictive capacity, while future work will evaluate additional specimens, and explore the sensitivity to uncertainties in experimental and modeling parameters. Acknowledgements. This work was supported by Orthosensor Inc. For any figures or tables, please contact authors directly (see Info & Metrics tab above).

Introduction. The longevity of total hip arthroplasty (THA) is dependent on acetabular component position. We measured the reliability and accuracy of a CT-based navigation system to achieve the intended acetabular component position and orientation using three dimensional imaging. The purpose of the current study was to determine if the CT-guided robotic navigation system could accurately achieve the desired acetabular component position (center of rotation (COR)) and orientation (inclination and anteversion). The postoperative orientation and location of the components was determined in 20 patients undergoing THA using CT images, the gold standard for acetabular component orientation. Methods. Twenty primary unilateral THA patients were enrolled in this IRB-approved, prospective cohort study to assess the accuracy of the robotic navigation system. Pre- and post-operative CT exams were obtained and aligned 3D segmented models were used to measure the difference in center of rotation and orientation (anteversion and inclination). Patients with pre-existing implants, posttraumatic arthritis, contralateral hip arthroplasty, septic arthritis, or previous hip fracture were excluded. All patients underwent unilateral THA using robotic arm CT-guided navigation (RIO Makoplasty; MAKO Surgical Corp). Results. Mean age was 59.25 years (±8.65 years), 55% of patients were female (11/20). Root mean square (RMS) errors between the intended intraoperative and actual postoperative COR position was measured in the medial/lateral (M/L), superior/inferior (S/I), and anterior/posterior (A/P) directions to quantify the accuracy of the CT-based robotic navigation system. The error in COR was variable (Fig. 4). The M/L distance error was 1.29 mm (SD: 1.18 mm; range: −2.61 – 1.13 mm). The S/I distance error was 1.81 mm (SD: 1.56 mm; range: −2.19 – 3.0 mm). The A/P distance error was 1.50 mm (SD: 1.50 mm; range: −3.53 – 2.23 mm). The mean difference between the intraoperative intended anteversion and postoperative actual anteversion was 2.2° ±1.6° with an RMS error of 2.73°. The mean difference in intraoperative intended inclination and postoperative actual inclination was 3.3° ± 1.7° with an RMS error of 3.71°. The robotic navigation system was more reliable in achieving the intended anteversion than intended inclination. The ICC for anteversion was 0.92 (95% CI 0.91–0.97), compared to ICC 0.74 (95% CI 0.49–0.89) for inclination. Conclusion. Our results suggest that CT-based navigation for THA is accurate for achieving intended cup center of rotation and both reliable and accurate in reproducing the intended cup orientation. Future research will focus on the use of a CT-based robotic navigation system to assist surgeons in the execution of a kinematic-based plan to eliminate impingement to reduce THA instability while maximizing range of motion

Introduction. Computer-assisted orthopaedic surgery (CAOS) has been shown to assist in achieving accurate and reproducible prosthesis position and alignment during total knee arthroplasty (TKA) [1]. The most prevalent modality of navigator tracking is optical tacking, which relies on clear line-of-sight (visibility) between the localizer and the instrumented trackers attached to the patient. During surgery, the trackers may not always be optimally positioned and orientated, sometimes forcing the surgeon to move the patient's leg or adjust the camera in order to maintain tracker visibility. Limited information is known about tracker visibility under clinical settings. This study quantified the rotational limits of the trackers in a contemporary CAOS system for maintaining visibility across the surgical field. Materials and Methods. A CAOS system (ExactechGPS®, Blue-Ortho, Grenoble, FR) was set up in an operating room by a standard surgical table according to the manufacture's recommendation. A grid with 10×10 cm sized cells was placed at the quadrant of the surgical table associated with the TKA surgical field [Fig. 1A,B]. The localizer was set up to aim at the center of the grid. A TKA surgical procedure was then initiated using the CAOS system. Once the trackers-localizer connection was established, the CAOS system constantly monitored the root mean square error (RMS) of each tracker. The connection was immediately aborted if the measured RMS was above the defined threshold. Therefore, “visibility” was defined as the tracker-localizer connection with proper accuracy level. An F tracker from the tracker set (3 trackers with similar characteristics) was placed at the center of each cell by a custom fixture, facing along the +Y axis [Fig. 1]. The minimum and maximum angles of rotation around the Z axis (RAZ_MIN and RAZ_MAX) and X axis (RAX_MIN and RAX_MAX) for maintaining tracker visibility were identified. For each cell, the rotational limit of the tracker was calculated for each axis of rotation as the difference between the maximum and minimum angles (RLX and RLZ). Results. The tracker rotation limits were 144.7±3.9° for RLZ (range: 136°–152°), and 150.5±3.9° for RLX (range: 143°–158°). RLX was significantly higher than RLZ across the field (difference in means=5.8°, p<0.01). Along the X axis, the rotational limit decreased slightly for RLZ, but increased slightly for RLX [Fig. 2]. Discussion. Studies have pointed out that the need for maintaining line-of-sight can be a limitation for the use of optical tracking based CAOS systems [2,3]. The results here demonstrated that ExactechGPS provides tracker visibility for more than 135° rotation across the surgical field. Moreover, the system is placed inside the sterile field, eliminating the potential blockage of the optical localizer by the surgical staff, further ensuring tracker visibility. The slight rotational limits trends along the X axis may be due to camera placement at one side of the surgical table. The current methodology may be applied to other CAOS systems to quantify the tracker visibility in a clinical environment. To view tables/figures, please contact authors directly

Are there any patho-anatomical features that might predispose to primary knee OA? We investigated the 3D geometry of the load bearing zones of both distal femur and proximal tibias, in varus, straight and valgus knees. We then correlated these findings with the location of wear patches measured intra-operatively. Patients presenting with knee pain were recruited following ethics approval and consent. Hips, knees and ankles were CT-ed. Straight and Rosenburg weight bearing X-Rays were obtained. Excluded were: Ahlbäck grade “>1”, previous fractures, bone surgery, deformities, and any known secondary causes of OA. 72 knees were eligible. 3D models were constructed using Mimics (Materialise Inc, Belgium) and femurs oriented to a standard reference frame. Femoral condyle Extension Facets (EF) were outlined with the aid of gaussian curvature analysis, then best-fit spheres attached to the Extension, as well as Flexion Facets(FF). Resected tibial plateaus from surgery were collected and photographed, and Matlab combined the average tibia plateau wear pattern. Of the 72 knees (N=72), the mean age was 58, SD=11. 38 were male and 34 female. The average hip-knee-ankle (HKA) angle was 1° varus (SD=4°). Knees were assigned into three groups: valgus, straight or varus based on HKA angle. Root Mean Square (RMS) errors of the medial and lateral extension spheres were 0.4mm and 0.2mm respectively. EF sphere radii measurements were validated with Bland-Altman Plots showing good intra- and interobserver reliability (+/− 1.96 SD). The radii (mm) of the extension spheres were standardised to the medial FF sphere. Radii for the standardised medial EF sphere were as follows; Valgus (M=44.74mm, SD=7.89, n=11), Straight (M=44.63mm, SD=7.23, n=38), Varus (M=50.46mm, SD=8.14, n=23). Ratios of the Medial: Lateral EF Spheres were calculated for the three groups: Valgus (M=1.35, SD=.25, n=11), Straight (M=1.38, SD=.23, n=38), Varus (M=1.6, SD=.38, n=23). Data was analysed with a MANOVA, ANOVA and Fisher's pairwise LSD in SPSS ver 22, reducing the chance of type 1 error. The varus knees extension facets were significantly flatter with a larger radius than the straight or valgus group (p=0.004 and p=0.033) respectively. In the axial view, the medial extension facet centers appear to overlie the tibial wear patch exactly, commonly in the antero-medial aspect of the medial tibial plateau. For the first time, we have characterised the extension facets of the femoral condyles reliably. Varus knees have a flatter medial EF even before the onset of bony attrition. A flatter EF might lead to menisci extrusion in full extension, and early menisci failure. In addition, the spherical centre of the EF exactly overlies the wear patch on the antero-medial portion of the tibia plateau, suggesting that a flatter medial extension facet may be causally related to the generation of early primary OA in varus knees

INTRODUCTION. Thorough understanding and feedback of the post-operative implant position relative to the pre-operative anatomy is missing in today's clinical practice. However, three dimensional insights in the local under or oversizing of the implant can provide important feedback to the surgeon. For the knee for instance, to identify a shift in the sagittal joint line that potentially links to mid-flexion instability or to identify zones at risk for soft tissue impingement. Despite a proven inferior outcome, clinical post-operative implant evaluation remains primarily based on bi-planar, static 2D x-rays rather than 3D imaging. Along with the cost, a possible reason is the increased radiation dose and/or metal artifact scatter in computed tomography (CT) and/or magnetic resonance imaging (MRI). These detrimental effects are now avoided by using recently released x-ray processing software. This technique uses standard-of-care post-operative x-rays in combination with a pre-operative CT and 3D file of the implant to determine the implant position relative to the pre-operative situation. The accuracy of this new technique is evaluated in this paper using patient cases. Therefore, the obtained implant position is benchmarked against post-operative CT scans. MATERIALS & METHODS. Retrospectively, 19 patients were selected who underwent total knee arthroplasty and received pre- and post-operative CT of their diseased knee. The CT scans were performed with a pixel size of 0.39 mm and slice spacing of 0.60 mm (Somatom, Siemens, München, Germany). All patients underwent TKA surgery using the same bi-cruciate substituting total knee (Journey II, Smith&Nephew, Memphis, USA). Following surgery, standard bi-planar standing x-rays of the operated knee was additionally performed as standard of care. To evaluate the implant position relative to the pre-operative situation, the 3D implants are first positioned on the post-operative CT slices. Using Mimics (Materialise NV, Leuven, Belgium), the pre-operative bone was subsequently automatically matched onto the post-operative scan to identify the implant location relative to the reconstructed pre-operative bone. This has been independently repeated by three observers to assess the inter-observer variability. Second, the post-operative bi-planar x-rays are combined with the reconstructed pre-operative bone and 3D file of the implant. This combination is performed using the 2D-to-3D conversion integrated in the recently launched X-ray module of Mimics. This module uses a contour based registration method to determine the implant and bone position using the post-operative x-rays. For both reconstruction methods, the implant position has been evaluated in six degrees of freedom using an automated Matlab routine; resulting in three translations and three rotations. RESULTS. From the evaluated implant positions, the root mean square error was derived between subsequent measurements. For the CT reconstruction based inter-observer evaluation, the median RMS error for all degrees of freedom is below 1 mm and 1 degree for both the femoral and tibial implant. Comparing the reconstructed CT implant position with the 2D-to-3D reconstruction, the median RMS difference between the implant positions remains below 1 mm and 1 degree except for the distraction/compression component and the internal/external rotation of the component. DISCUSSION. On average, the RMS difference between the 2D-to-3D conversion and the reconstructed post-operative CT exceeds the inter-observer RMS difference obtained using reconstructed post-operative CT. The differences are in line with previous cadaveric studies using the same reconstruction technique. The largest differences are seen for the femoral and tibial internal/external rotation. However, the obtained values are still within reasonable limits according to a recent review by De Valk et al., who reported an inter-observer variation of 3° for the femur and 2° for the tibia. In addition, the 2D-to-3D conversion displays a larger difference for the distraction/compression component. Since a true, golden standard measurement is lacking in our tests, it is not clear whether this error is attributed to the CT imaging or the 2D-to-3D conversion. Given the low inter-observer variation for this degree of freedom, it is hypothesized that this discrepancy is linked to the finite slice spacing for the CT scans. Apart from the obtained accuracy, the use of the 2D-to-3D module has the advantage of significantly reducing the radiation dose with approx. a factor 20. In addition, the imaging procedure needs no more than the standard imaging required by clinical practice

Background. Total hip arthroplasty (THA) is one of the most successful surgical procedures ever performed. Nevertheless if procedure is performed by high or low volume surgeons; more than 50% of cups are still placed out of the safe zone, which is connected to lower survival rate of the prosthesis. The idea was to develop an imageless navigation system for safe and accurate positioning of the cup in THA procedures, without a need of any preoperative computer tomography (CT) or magnetic resonance imagining (MRI). Methods. The validation of the system was approved by National Ethics Committee. The committee allowed the validation on 10 patients who all signed the agreement for participation in the study. Unselected patients undergoing THA were included. All patients had had performed preoperative x-rays of pelvis and hips for standard preoperative planning. Immediately before skin incision, anterior pelvic plane (APP) was defined with help of specially developed electromagnetic navigation system (Guiding Star, E-Hip module, Ekliptik d.o.o., Ljubljana, Slovenia) and specificaly designed hardware tool which is essential for accurate APP determination [Fig.1]. In all patients THAs were performed through direct lateral approach and all implanted components (Allofit S cup and Alloclassic stem, Zimmer Inc., Warsaw, Indiana, USA) were implanted with freehand technique according to preoperative plan. After placement of the cups their inclination and anteversion angles were determined with aforementioned navigation system [Fig. 2]. The day after surgery, low dose CT scans of pelvises of operated patients were performed and DICOM format files were up-loaded into EBS software (Ekliptik d.o.o., Ljubljana, Slovenia), a multipurpose application for perioperative planning, measuring and constructing where virtual copies of pelvises were generated. On virtual pelvises the position of the cups was measured by independent person [Fig.3]. Measurements were compared, statistically analysed and the deviation calculated with root mean square error (RMSE) method. Afterwards the average error (eaver) and standard deviation (σ) between intraoperatively determined and postoperatively measured angles were calculated. Results. We included 10 patients in the study, with 6 left and 4 right hips. The maximal and minimal differences between navigation and CT measurements for inclination angles were 5.3° and 0.3° respectively, with calculated eaver of 0.7°, σ of 2.6° and RMSE of 2.6°. The maximal and minimal differences between navigation and CT measurements for anteversion angles were 4.6° and 0.7° respectively, with calculated eaver of −1.9°, σ of 1.8° and RMSE of 2.6°. Conclusion. We determined that the imageless navigation system we validated is a very accurate tool for cup placement in THA. The accuracy of the system is within 2° which by far exceeds the abilities of the best freehand techniques. In line to the trends, supporting more precise and less invasive surgery, the THA with help of imageless navigation should in our opinion become a golden standard, especially in minimally invasive procedures

INTRODUCTION. The purpose of this study is to elucidate longitudinal kinematic changes of the hip joint during heels-down squatting after THA. METHODS. 66 patients with 76 primary cementless THAs using a CT-based navigation system were investigated using fluoroscopy. An acetabular component and an anatomical femoral component were used through the mini-posterior approach with repair of the short rotators. The femoral head size was 28mm (9 hips), 32mm (12 hips), 36mm (42 hips), and 40mm (12 hips). Longitudinal evaluation was performed at 3 months, 1 year, and 2≤ years postoperatively. Successive hip motion during heels-down squatting was recorded as serial digital radiographic images in a DICOM format using a flat panel detector. The coordinate system of the acetabular and femoral components based on the neutral standing position was defined. The images of the hip joint were matched to 3D-CAD models of the components using a2D/3D registration technique. In this system, the root mean square errors of rotation was less than 1.3°, and that of translation was less than 2.3 mm. We estimated changes in the relative angle of the femoral component to the acetabular component, which represented the hip ROM, and investigated the incidence of bony and/or prosthetic impingement during squatting (Fig.1). We also estimated changes in the pelvic posterior tilting angle (PA) using the acetabular component position change. In addition, when both components were positioned most closely during squatting, we estimated the minimum angle (MA) up to theoretical prosthetic impingement as the safety margin (Fig.2). RESULTS. No prosthetic or bony impingement and no dislocation occurred in any hips. The mean maximum hip flexion ROM was 92.4° (range, 76.6° – 107.9°) at 3 months, 103.4° (range, 81.5° – 115.2°) at 1 year, and 102.4° (range, 87.1° – 120.6°) at 2≤ years (3 months vs 1 year, p<0.05; 1 year vs 2≤ years, p>0.05, paired t-test). The mean PA was 26.7° (range, 0.9° – 49.8°) at 3 months, 21.7° (range, 3.4° – 43.8°) at 1 year, and 21.2° (range, −0.7° – 40.4°) at 2≤ years (3 months vs 1 year, p<0.05; 1 year vs 2≤ years, p>0.05). The mean flexion ROM and MA at 2≤ years were 98.4±20.8° and 14.3±7.3° in 28 mm heads, 102.3±10.7° and 15.6±4.8° in 32 mm heads, 102.8±14.5° and 20.3±9.6° in 36 mm heads, and 103.2±16.9° and 23.4±10.9° in 40 mm heads, respectively. There were no significant differences in the hip flexion ROM between 28, 32, 36, and 40 mm head cases, whereas MA significantly increased as the femoral head diameter was larger (p<0.05, unpaired t-test). DISCUSSION AND CONCLUSION. Three-dimensional assessment of dynamic squatting motion after THA using the 2D/3D registration technique enabled us to elucidate longitudinal kinematic change of the hip joint. Longitudinal kinematic analysis indicated that hip flexion ROM and posterior tilt during squatting changed significantly by 1 year postoperatively, and there were no significant changes after 1 year while safety margin kept > 10°. For figures/tables, please contact authors directly.

INTRODUCTION. Understanding the relationship between knee specific tissue behavior and joint contact mechanics remains an area of focus. Seminal work from 1990's established the possibility to optimize tissue properties for recreation of laxity driven kinematics (Mommersteeg et al., 1996). Yet, the uniqueness and validity of such predictions could be strengthened, especially as they relate to joint contact conditions. Understanding this interplay has implications for the long term performance of joint replacements. Development of instrumented knee implants, highlighted by a single use tibial insert trial with embedded sensor technology (VERASENSE, Orthosensor Inc.), may offer an avenue to establish the relationship between tissue state and joint mechanics. Utilization of related data also has the potential to confirm computational predictions, where both rigid body motions and associated reactions are explicitly accounted for. Hence, the goal of this work was to evaluate an approach for optimization of ligament properties using joint mechanics data from an instrumented implant during laxity style testing. Such a framework could be used to inform joint balancing techniques, improve long term implant performance, and alternatively, qualify factors that may lead to poor outcomes. METHODS. Experimentation was performed on a 52 year old male, left, cadaveric specimen. Joint arthroplasty was performed using standard practice by an experienced orthopedic surgeon. To mimic passive intraoperative loading, laxity loading at 10°, 45° and 90° flexion, which consisted of discrete application of anterior-posterior (± 100N), varus-valgus (± 5 Nm) and internal-external (± 3 Nm) loads at each angle, was performed using a simVITROTM robotic musculoskeletal simulator (Cleveland Clinic, Cleveland, OH). Experimental results included relative tibiofemoral kinematics and sensor measured metrics (Fig 1). The finite element model was developed from specimen-specific MRIs and solved using Abaqus/Explicit. The model included the rigid bones, appropriately placed implants and relevant soft-tissue structures (Fig. 1). Ligament stiffness values were adopted from the literature and included a 6% strain toe region. Sets of nonlinear springs, defined using MR imaging, comprised each ligament/bundle. Optimization was performed, which minimized the root mean squared difference between VERASENSE measured tibiofemoral mechanics and the model predicted values. Ligament slack lengths were the control variables and the objective included each loading state and all contact metrics (θ, AFD, ML, and LL). RESULTS AND DISCUSSION. The model successfully recreated joint kinematics with average errors of 4° for rotations and 3 mm for translations, across all flexion angles (Fig 2). Though a systematic offset in θ was observed, model versus experiment contact locations were also in good agreement. Reaction forces were generally over-predicted by the model, but retained the overall trend (Fig 2). Sensitivity analysis also supported this finding. In light of the larger focus of this project, testing also included systematic removal of key tissues followed by repeat testing, as evaluated across numerous specimens. Overall, the presented framework represents a promising step towards establishing simulation based tools able to support exploratory studies as well as the clinical decision making process. Future work will evaluate efficacy across numerous specimens and assess sensitivity to key modeling and experimental parameters. To view tables/figures, please contact authors directly