Total Knee Arthroplasty Kinematics Predict Patient-Reported Outcome Measures: Implications for Clinical Kinematic Examinations.

BACKGROUND: A core tenet of total knee arthroplasty (TKA) is that achieving more natural kinematics will lead to superior patient outcomes. Yet this relationship has not been proven for large representative cohorts of TKA patients because accurately measuring 3-dimensional TKA kinematics is time-consuming and expensive. But advanced imaging systems and machine learning-enhanced analysis software will soon make it practical to measure knee kinematics preoperatively and postoperatively in the clinic using radiographic methods. The purpose of this study was to assess the reported relationships between TKA kinematics and outcomes and distill those findings into a proposal for a clinically practical protocol for a clinical kinematic exam. METHODS: This study reviewed the recent literature relating TKA kinematics to patient outcomes. There were 10 studies that reported statistical associations between TKA kinematics and patient outcome scores utilizing a range of functional activities. We stratified these activities by the complexity of the radiographic examination to create a proposed examination protocol, and we generated a list of requirements and characteristics for a practical TKA clinical kinematic examination. RESULTS: Given considerations for a clinically practical kinematic exam, including equipment, time and other resources, we propose 3 exam levels. With basic radiographs, we suggest studying single-leg stance in extension, lunge or squat, and kneeling. For fluoroscopic systems with X-ray pulses up to 20 ms, we propose chair-rise or stair ascent to provide additional dynamic information. For fluoroscopic systems with X-ray pulses of less than 10 ms, we propose rapid open-chain knee flexion-extension to simulate the highly dynamic swing phase of gait. CONCLUSIONS: It is our hope that this proposed examination protocol spurs discussion and debate so that there can be a consensus approach to clinical examination of knee and TKA kinematics when the rapidly advancing hardware and software capabilities are in place to do so.

Soft-tissue tension during total hip arthroplasty measured in four patients and predicted using a musculoskeletal model.

PURPOSE: Soft-tissue tension around the hip joint is related to the incidence of dislocation after total hip arthroplasty (THA), but it remains difficult to quantify the soft-tissue tension during surgery. In this study, a three-dimensional force sensor-instrumented modular femoral head was developed and used to quantify soft-tissue tension during THA. The forces at the hip joint were also calculated using a three-dimensional musculoskeletal computer model to validate the measured forces. METHODS: Soft-tissue tension was investigated by measuring the hip joint forces and directions during intraoperative trialing in four patients through passive range of motion (ROM) from 0° extension to 90° flexion. A musculoskeletal model with THA, which was scaled to one of four patients, was developed. The hip joint forces were calculated under the same motion. RESULTS: Through the passive ROM, the magnitude of soft-tissue tension was greatest when the hip was extended, decreased with flexion to 34°, and progressively increased to flexion at 90°. The mediolateral force component was relatively constant, but the supero-inferior and anterior-posterior force components changed significantly. Within-individual variations were small during three repeated cycles of measurement, but magnitudes varied significantly among patients. Similar force patterns and magnitudes were calculated by the musculoskeletal model. CONCLUSIONS: This study demonstrates that it is possible to quantify soft-tissue tension and direction during THA with an instrumented head. There was general agreement between the calculated and measured forces in both pattern and magnitude. Including additional subject-specific details would further enhance agreement between the model and measured hip forces.

In Vivo Kinematic Analysis of the Axial Shoulder Rotation in the Standing and Supine Positions Using 3D/2D Registration and Electromyography.

Background There has been no report comparing shoulder kinematics and muscle activities during axial shoulder rotation in different positions. The purpose of this study was to investigate differences in shoulder kinematics and muscle activities during axial shoulder rotation in healthy subjects between standing and supine positions using three-dimensional/two-dimensional (3D/2D) registration techniques and electromyography (EMG). Methods Eleven healthy males agreed to participate in this study. We recorded the fluoroscopy time during active shoulder axial rotation with a 90° elbow flexion in both standing and supine positions, simultaneously recording surface EMG of the infraspinatus, anterior deltoid, posterior deltoid, and biceps brachii. Three-dimensional bone models were created from CT images, and shoulder kinematics were analyzed using 3D/2D registration techniques. Muscle activities were evaluated as a ratio of mean electromyographic values to 5-sec maximum voluntary isometric contractions.  Results Scapular kinematics during axial shoulder rotation in the supine position showed similar patterns with those in the standing position. The scapula was more posteriorly tilted and more downwardly rotated in the supine posture than in standing (P < 0.001 for both). Acromiohumeral distance (AHD) in the supine posture was significantly larger than in standing. Muscle activities showed no significant differences between postures except for biceps (P < 0.001). Discussion Shoulder kinematics and muscle activities during axial rotation were similar in pattern between standing and supine postures, but there were shifts in scapular pose and AHD. The findings of this study suggest that posture may be an important consideration for the prescription of optimal shoulder therapy following surgery or for the treatment of shoulder disorders.

DRR acceleration using inexpensive GPUs for model-image registration based joint kinematic measurements.

Model-image registration methods are commonly used in research to measure three-dimensional joint kinematics from single-plane and bi-plane x-ray images. These methods have the potential to be beneficial if used clinically, but current techniques are too slow or expensive to be clinically practical. One technical element of these methods for measuring natural bone motion is the use of digitally reconstructed radiographs (DRRs). DRRs can be very expensive to compute, or require expensive and fast computer hardware. In this technical development, a numerically efficient Siddon-Jacobs algorithm for computing DRRs was implemented on a consumer-grade graphics card using a programming language for parallel architectures. Compared to traditional voxel projection algorithms with a central-processing-unit-only implementation, the parallel computation implementation on the graphics card provided speedups of 650-1546 times faster rendering, while retaining equivalent performance for joint kinematics measurements. The use of consumer grade graphics hardware may contribute to making model-image registration measurements of joint kinematics practical for clinical use.

An assessment of the fixin tplo jig to generate effective compression using a transverse fracture model.

The objective of this study was to determine compressive loads that could be generated using a tibial plateau leveling osteotomy (TPLO) jig with a tensioned strand of 18-gauge stainless steel orthopedic wire in a simulated transverse fracture model. The wire was sequentially tensioned using heavy needle holders or an AO wire tightener. Recorded loads were subsequently compared to loads generated by applying a 3.5 mm limited contact-dynamic compression plate (LC-DCP) as a compression plate. Two segments of 2 cm diameter Delrin rod were placed in a testing apparatus and used to simulate a transverse fracture. A load cell was interposed between the two segments to measure the compressive loads generated during the application of the TPLO jig or the LC-DCP. Compression was generated by sequential tensioning a strand of 18-gauge wire secured through the base of the arms of the TPLO jig or by placing one or two load screws in the LC-DCP. Wires were tensioned using heavy needle holders or an AO wire tightener. Eight replicates of each construct were tested. Recorded loads were compared using a one-way repeated measures ANOVA and Tukey Honestly Significant Difference test. The wire being tensioned broke while attempting a second quarter rotation of the needle holders and when the crank handle of the AO wire tightener was advanced beyond two rotations. The mean + SD peak compressive loads recorded when tensioning the wire using the heavy needle holders and AO wire tightener was 148 ± 7 N and 217 ± 16 N, respectfully. The mean ± SD load recorded after placement of the first and second load screw in the LC-DCP was 131 ± 39 N and 296 ± 49 N, respectively. The compression generated by placing two load screws in the LC-DCP was superior to the compression generated using the jig. The maximum load recorded by tensioning the wire secured through the TPLO jig using the AO wire tightener was superior to the compression generated by placing a single load screw and tensioning the wire using needle holders. Our results demonstrate that the TPLO jig allows surgeons to compress transverse fractures or osteotomies effectively. Tensioning the AO wire tightener allows for sequential tensioning and generates superior compressive loads than tensioning wires with heavy needle holders.

Anatomical variations of the equine femur and tibia using statistical shape modeling.

The objective of this study was to provide an overarching description of the inter-subject variability of the equine femur and tibia morphology using statistical shape modeling. Fifteen femora and fourteen tibiae were used for building the femur and tibia statistical shape models, respectively. Geometric variations in each mode were explained by biometrics measured on ±3 standard deviation instances generated by the shape models. Approximately 95% of shape variations within the population were described by 6 and 3 modes in the femur and tibia shape models, respectively. In the femur shape model, the first mode of variation was scaling, followed by notable variation in the femoral mechanical-anatomical angle and femoral neck angle in mode 2. Orientation of the femoral trochlear tubercle and femoral version angle were described in mode 3 and mode 4, respectively. In the tibia shape model, the main mode of variation was also scaling. In mode 2 and mode 3, the angles of the coronal tibial plateau and the medial and lateral caudal tibial slope were described, showing the lateral caudal tibial slope angle being significantly larger than the medial. The presented femur and tibia shape models with quantified biometrics, such as femoral version angle and posterior tibial slope, could serve as a baseline for future investigations on correlation between the equine stifle morphology and joint disorders due to altered biomechanics, as well as facilitate the development of novel surgical treatment and implant design. By generating instances matching patient-specific femorotibial joint anatomy with radiographs, the shape model could assist virtual surgical planning and provide clinicians with opportunities to practice on 3D printed models.

Joint Track Machine Learning: An Autonomous Method of Measuring Total Knee Arthroplasty Kinematics From Single-Plane X-Ray Images.

BACKGROUND: Dynamic radiographic measurements of 3-dimensional (3-D) total knee arthroplasty (TKA) kinematics have provided important information for implant design and surgical technique for over 30 years. However, current methods of measuring TKA kinematics are too cumbersome, inaccurate, or time-consuming for practical clinical application. Even state-of-the-art techniques require human-supervision to obtain clinically reliable kinematics. Eliminating human supervision could potentially make this technology practical for clinical use. METHODS: We demonstrate a fully autonomous pipeline for quantifying 3D-TKA kinematics from single-plane radiographic imaging. First, a convolutional neural network (CNN) segmented the femoral and tibial implants from the image. Second, those segmented images were compared to precomputed shape libraries for initial pose estimates. Lastly, a numerical optimization routine aligned 3D implant contours and fluoroscopic images to obtain the final implant poses. RESULTS: The autonomous technique reliably produces kinematic measurements comparable to human-supervised measures, with root-mean-squared differences of less than 0.7 mm and 4° for our test data, and 0.8 mm and 1.7° for external validation studies. CONCLUSION: A fully autonomous method to measure 3D-TKA kinematics from single-plane radiographic images produces results equivalent to a human-supervised method, and may soon make it practical to perform these measurements in a clinical setting.

A silicone-based support material eliminates interfacial instabilities in 3D silicone printing.

Among the diverse areas of 3D printing, high-quality silicone printing is one of the least available and most restrictive. However, silicone-based components are integral to numerous advanced technologies and everyday consumer products. We developed a silicone 3D printing technique that produces precise, accurate, strong, and functional structures made from several commercially available silicone formulations. To achieve this level of performance, we developed a support material made from a silicone oil emulsion. This material exhibits negligible interfacial tension against silicone-based inks, eliminating the disruptive forces that often drive printed silicone features to deform and break apart. The versatility of this approach enables the use of established silicone formulations in fabricating complex structures and features as small as 8 micrometers in diameter.

Validation of a machine learning technique for segmentation and pose estimation in single plane fluoroscopy.

Kinematics of total knee replacements (TKR) play an important role in assessing the success of a procedure and would be a valuable addition to clinical practice; however, measuring TKR kinematics is time consuming and labour intensive. Recently, an automatic single-plane fluoroscopic method utilizing machine learning has been developed to facilitate a quick and simple process for measuring TKR kinematics. This study aimed to validate the new automatic single-plane technique using biplanar radiostereometric analysis (RSA) as the gold standard. Twenty-four knees were imaged at various angles of flexion in a dedicated RSA lab and 113 image pairs were obtained. Only the lateral RSA images were used for the automatic single-plane technique to simulate single-plane fluoroscopy. Two networks helped automate the kinematics measurement process, one segmented implant components and the other generated an initial pose estimate for the optimization algorithm. Kinematics obtained via the automatic single plane and manual biplane techniques were compared using root-mean-square error and Bland-Altman plots. Two observers measured the kinematics using the automated technique and results were compared with assess reproducibility. Root-mean-square errors were 0.8 mm for anterior-posterior translation, 0.5 mm for superior-inferior translation, 2.6 mm for medial-lateral translation, 1.0° for flexion-extension, 1.2° for abduction-adduction, and 1.7° for internal-external rotation. Reproducibility, reported as root-mean-square errors between operator measurements, was submillimeter for in-plane translations and below 2° for all rotations. Clinical Significance: The advantages of the automated single plane technique should aid in the kinematic measurement process and help researchers and clinicians perform TKR kinematic analyses.

Quantification of equine stifle passive kinematics.

OBJECTIVE: This study aims to quantitatively characterize the passive kinematics of the healthy, soft tissue-intact equine stifle to establish an objective foundation for providing insights into the etiology of stifle disorders and developing a possible surgical treatment for stifle degenerative disease. ANIMALS: 5 whole-horse specimens. PROCEDURES: Reflective markers with intracortical bone pins and a motion capture system were used to investigate the stifle joint kinematics. Kinematics of 5 whole-horse specimens euthanized within 2 hours were calculated for internal/external rotation, adduction/abduction, and cranial/caudal translation of the medial and lateral femoral condyles and estimated joint contact centroids as functions of joint extension angle. RESULTS: From 41.7° to 121.6° (mean ± SD, range of motion: 107.5° ± 7.2°) of joint extension, 13° ± 3.7° of tibial external rotation and 6° ± 2.7° of adduction were observed. The lateral femoral condyle demonstrated significantly greater cranial translation than the medial during extension (23.7 mm ± 9.3 mm vs. 14.3 mm ± 7.0 mm, P = .01). No significant difference was found between the cranial/caudal translation of estimated joint contact centroids in the medial and lateral compartment (13.3 mm ± 7.7 mm vs. 16.4 mm ± 5.8 mm, P = .16). CLINICAL RELEVANCE: The findings share similarities with kinematics for human knees and sheep and dog stifles, suggesting it may be possible to translate what has been learned in human arthroplasty to treatment for equine stifles.

Posterior cruciate-retaining total knee arthroplasty exhibits small kinematic changes in the first postoperative year.

PURPOSE: Fluoroscopic knee kinematics have historically been quantified at least 1 year after total knee arthroplasty (TKA). The purpose of this study was to longitudinally assess knee kinematics at 6-12 weeks, 6 months, and 1 year after TKA to determine if earlier evaluation may be justified. METHODS: Twenty-one patients participated after undergoing TKA with a posterior cruciate ligament-retaining fixed-bearing prosthesis. Fluoroscopic examinations of lunge, kneel, and step-up activities were performed at 12 ± 4 weeks (V1), 7 ± 2 months (V2), and 13 ± 2 months (V3) postoperatively. Images were analyzed using a three-dimensional to two-dimensional image registration technique. Maximum flexion poses for lunging and kneeling were compared between visits with repeated-measures statistical tests. For the step-up activity, mixed-effects linear models were constructed for condylar anteroposterior (AP) contact points and tibial internal rotation throughout flexion. Estimated marginal means of fitted values were plotted with 95% confidence intervals and used to compare mean kinematics between visits. RESULTS: There were no significant changes in maximum lunging flexion over time (p = 0.405), though significant increases in maximum kneeling flexion were observed between V1 (106 ± 8°) and V2 (110 ± 9°) (p = 0.006), and V1 and V3 (113 ± 9°) (p = 0.0003). While statistical differences were calculated for lunging medial condyle AP translation and kneeling tibial internal rotation, absolute differences in condylar AP contact locations were less than ~ 2 mm between all visits during both movements. For the step-up activity, tibial internal rotation increased with flexion, and there were pair-wise significant differences at all flexion angles between V1-V2 (p < 0.001) and V1-V3 (p < 0.001). Anterior medial condylar translation was observed with flexion, with pair-wise significant differences present for V1-V3 (p = 0.005) and V2-V3 (p < 0.001). The lateral condyle exhibited initial posterior translation followed by anterior translation with increasing flexion, with pair-wise differences between all visits (p < 0.005 for all comparisons). CONCLUSION: Though statistical differences were observed between visits for all activities, variations in estimated mean condylar positions were within ~ 2 mm from ~ 12 weeks to 1 year. Considering measurement error averages approximately 1 mm for sagittal plane translations, these results indicate that knee kinematics during kneel, lunge, and step-up activities may be sustained from as early as 12 weeks after TKA.

Stifle kinematics in 4 dogs with cranial cruciate ligament insufficiency treated by CORA-based leveling osteotomy.

Objective: The purpose of this study was to quantify three-dimensional (3D) stifle kinematics during walking in dogs with complete cranial cruciate ligament insufficiency (CCL-I) treated with a CORA-based leveling osteotomy (CBLO). Study design: Four client-owned dogs with unilateral complete CCL-I were prospectively enrolled. Custom digital 3D models of the femora and tibiae were created from pre-and postoperative computed tomographic scans for each dog. Lateral view fluoroscopic images were collected during treadmill walking preoperatively and 6 months after CBLO. Results were generated using a 3D-to-2D image registration process. Pre-and postoperative stifle kinematics (craniocaudal translation, extension angle) were compared to that of the unaffected contralateral (control) stifle. Force plate gait analysis was performed, and symmetry indices (SI) were calculated for peak vertical force (PVF) and vertical impulse (VI). Results: After CBLO, craniocaudal femorotibial motion was reduced by a median (range) of 43.0 (17.0-52.6) % over the complete gait cycle. Median (range) PVF SI was 0.49 (0.26-0.56) preoperatively and 0.92 (0.86-1.00) postoperatively, and VI SI was 0.44 (0.20-0.48) preoperatively and 0.92 (0.82-0.99) postoperatively. Conclusion: CBLO mitigated but did not fully resolve abnormal craniocaudal translation; lameness was substantially improved at 6 months.

Kinematics of the equine distal sesamoid (navicular) bone of the thoracic limb.

OBJECTIVE: To quantify the translation and angular rotation of the distal sesamoid bone (DSB) using computed tomography (CT) and medical modeling software. SAMPLE: 30 thoracic limbs from equine cadavers. PROCEDURES: Partial (n = 12), full (8), and matched full and subsequently transected (10) thoracic limbs were collected. Bone volume CT images were acquired in three positions: extension (200° metacarpophalangeal angle), neutral (180°), and maximal flexion (110°). Mean translation and angular rotation of each DSB were recorded. Differences were determined with two-way ANOVA and post hoc Tukey's tests for pairwise comparisons; P value was set at < 0.05. RESULTS: Dorsal translation was significant during extension (1.4 ± 0.4 mm full limbs and 1.3 ± 0.2 mm partial limbs, P < 0.001). Distal translation was significant during extension (1.9 ± 0.4 mm full and 1.1 ± 0.4 mm partial) and flexion (5.4 ± 0.7 mm full and 6.22 ± 0.6 mm partial, P < 0.001). Rotation was significant (P < 0.001) about the mediolateral axis during extension (17.1° ± 1.4°) and flexion (2.6° ± 1.3°). Translation and rotation of the DSB were significantly different (P < 0.001) between full and partial limbs. CLINICAL RELEVANCE: This study provides the first quantification of translation and angular rotation of the DSB within the equine hoof. Partial limbs had significantly reduced movement compared to full limbs, suggesting that transection of flexor tendons alters distal thoracic limb kinematics. Further studies are required to determine if pathologic changes in the podotrochlear apparatus have an impact in clinical lameness outcomes.

Smaller femoral neck anteversion in varus knees than in healthy and valgus knees.

It is important to investigate anatomical differences of the femur and tibia three-dimensionally between varus and valgus knees to enhance surgical approaches and better understand structural factors related to specific patterns of osteoarthritis progression. Three-dimensional femoral and tibial bone models were reconstructed from transverse computed-tomography scans in varus osteoarthritis (43 knees), valgus osteoarthritis (40 knees), and healthy (32 knees) groups. Different coordinate systems were defined in each femoral bone model ("Knee" and "Hip" Coordinate System). Femoral neck inclination, lateral bowing, anterior bowing, and neck anteversion were measured and compared between knee and hip coordinate systems. Those parameters were also compared between varus, valgus, and healthy groups. The tibial anterior and lateral bowing, external torsion and the fibular axis relative to the tibial mechanical axis were measured and compared between varus and valgus groups. Femoral neck anteversion was significantly 1-2° greater in the hip coordinate system compared to the knee coordinate system. Femoral neck anteversion was significantly smaller in varus knees than in healthy or valgus knees, with average difference of approximately 5°. The knee and hip joint are often rotated externally relative to the trunk axis in patients with varus osteoarthritis, perhaps maintaining the geometric relations between pelvis and proximal femur (including peripheral hip muscles) regardless of knee deformities. The fibular axis was inclined slightly valgus and posteriorly in two groups. The results may inform hypotheses on, and future studies of, skeletal morphologic development and factors contributing to the progression of knee osteoarthritis.

The Effect of Posterior Cruciate Ligament Release on Kinematics and Outcomes in Primary Total Knee Arthroplasty With a Dual-Pivot Conforming Polyethylene.

BACKGROUND: Ultracongruent bearings are increasingly utilized in total knee arthroplasty (TKA); however, implications of surgical technique on knee kinematics and outcomes with these bearings are not well understood. This study's purpose was to evaluate the relationship of 3-dimensional knee kinematics and patient-reported outcome measures (PROMs) in a dual-pivot congruent bearing TKA with and without posterior cruciate ligament (PCL) release. METHODS: Forty patients undergoing TKA with an asymmetric ultracongruent bearing were prospectively enrolled for gait analysis preoperatively and 4 months postoperatively. Three-dimensional gait analysis was performed utilizing infrared motion capture. Knee kinematic data and PCL disposition were analyzed for correlations with PROMs. RESULTS: The PCL was fully released in 52.5% (21/40) of cases. Greater maximum anteroposterior femoral translation correlated with lower Knee Injury and Osteoarthritis Outcome Score for Joint Replacement (rho = -0.596, P = .012), greater Knee Society Score pain with level walking (rho = 0.411, P = .101), and greater Knee Society Score pain while climbing stairs (rho = 0.469, P = .058) at 4-month follow-up. The PCL-release group was associated with greater maximum femoral anteroposterior translation (9.8 vs 5.5 mm, P = .053) and greater maximum internal tibial rotation (-6.2° vs -3.0°, P = .040), supporting a more anterior-based position of the medial condyle. The PCL released group had lower median Knee Injury and Osteoarthritis Outcome Score for Joint Replacement scores (70.7 vs 76.3, P = .031) and reported that their knees "sometimes or always" feel normal less frequently (81.8% vs 92.3%, P = .576). CONCLUSION: With this asymmetric ultracongruent bearing TKA, preservation or partial titration release of the PCL, as opposed to full PCL release, appears to minimize deleterious anterior femoral translation and internal tibial rotation, which is correlated with optimized patient-reported outcomes. LEVEL OF EVIDENCE: II.

Comparing in vivo three-dimensional shoulder elevation kinematics between standing and supine postures.

BACKGROUND: It is often assumed that body posture, standing vs. supine, changes shoulder muscle activation and range of motion, but these altered shoulder mechanics have not been objectively assessed. We expected the supine posture might facilitate scapular rotation and change subacromial pressure. The purpose of this study is to evaluate the influence of body posture on shoulder kinematics during arm elevation. METHODS: Ten males and eight females with a mean age of 33 years participated in this study. Shoulder kinematics were assessed during scapular plane elevation in the standing and supine postures by using single-plane fluoroscopic images. Kinematics were measured using 3-dimensional to 2-dimensional model-image registration techniques: matching the 3-dimensional bone model derived from computed tomography onto each fluoroscopic image. Glenohumeral superior/inferior translation, acromiohumeral distance, and scapular rotations were compared between the postures. The effect of sex also was evaluated. RESULTS: With the arm at the side position, the humeral head in the supine posture was located 0.5 mm superior compared to the standing posture (P < .001). During humeral elevation, the humeral head significantly shifted more inferiorly in the supine posture than in standing; the biggest mean difference was 0.6 mm, P = .003. But acromiohumeral distance during elevation was not significantly affected by the body posture (P = .05). Scapular upward rotation and posterior tilt were significantly different between the postures (P < .001). Sex had statistically significant, but quantitatively small, effects on shoulder kinematics. CONCLUSIONS: Body postures affect shoulder kinematics during humeral elevation. This knowledge will be useful to optimize rehabilitation exercises and for diagnostic insight.

Correction: Three-dimensional-printed custom guides for bipolar coxofemoral osteochondral allograft in dogs.

[This corrects the article DOI: 10.1371/journal.pone.0244208.].

Automatic tracking of healthy joint kinematics from stereo-radiography sequences.

Kinematic tracking of healthy joints in radiography sequences is frequently performed by maximizing similarities between computed perspective projections of 3D computer models and corresponding objects' appearances in radiographic images. Significant human effort associated with manual tracking presents a major bottleneck in biomechanics research methods and limits the scale of target applications. The current work introduces a method for fully-automatic tracking of tibiofemoral and patellofemoral kinematics in stereo-radiography sequences for subjects performing dynamic activities. The proposed method involves the application of convolutional neural networks for annotating radiographs and a multi-stage optimization pipeline for estimating bone pose based on information provided by neural net predictions. Predicted kinematics are evaluated by comparing against manually-tracked trends across 20 distinct trials. Median absolute differences below 1.5 millimeters or degrees for 6 tibiofemoral and 3 patellofemoral degrees of freedom demonstrate the utility of our approach, which improves upon previous semi-automatic methods by enabling end-to-end automation. Implementation of a fully-automatic pipeline for kinematic tracking will benefit evaluation of human movement by enabling large-scale studies of healthy knee kinematics.

Three-dimensional kinematics of reverse shoulder arthroplasty: a comparison between shoulders with good or poor elevation.

BACKGROUND: Various factors may be related to outcomes of reverse shoulder arthroplasty (RSA) including patient and surgical factors. Differences in shoulder kinematics might be associated with poor function after RSA; however, kinematic differences between shoulders with good or poor elevation have not been elucidated. The purpose of this study was to compare RSA kinematics between shoulders with good or poor elevation. METHODS: The study included 28 shoulders with a minimum 6-month follow-up after RSA using Grammont-type prostheses. Subjects comprised 17 men and 11 women with the mean age of 75 years (range, 63-91). Subjects underwent fluoroscopy during active scapular plane abduction. Computed tomography of their shoulders was performed to create 3-dimensional scapular implant models. Using model-image registration techniques, poses of 3-dimensional implant models were iteratively adjusted to match their silhouettes with the silhouettes in the fluoroscopic images, and 3-dimensional kinematics of implants were computed. Kinematics and glenosphere orientation were compared between shoulders with good (>90 degree) or poor (<90 degree) scapular plane abduction. RESULTS: Nineteen and 9 shoulders were assigned to the good- and poor-elevation groups, respectively. There were no significant differences between the groups in age, sex, height, weight, preoperative range of motion, or Constant score, but body mass index in the poor elevation shoulders was significantly larger than that in the good elevation shoulders. There were no significant differences in glenosphere (upward/downward rotation, anterior/posterior tilt, internal/external rotation) or glenohumeral (internal/external rotation, abduction/adduction) kinematics between the good and poor elevation shoulders. Scapulohumeral rhythm was significantly higher in the good elevation shoulders than the poor elevation shoulders (P = .04). Glenosphere superior tilt was 2.3° ± 4.2° in the good-elevation group and 8.1° ± 8.9° in the poor-elevation group, and the difference was statistically significant (P = .03). DISCUSSION: Shoulders with good elevation after RSA demonstrated better scapulohumeral rhythm than those with poor elevation, though there were no significant differences in glenosphere and glenohumeral kinematics. It may be important for better elevation to achieve good glenohumeral motion in shoulders with RSA. Glenosphere orientations may affect postoperative shoulder function.