Comparison of the early postoperative outcomes of cementless and cemented medial unicompartmental knee arthroplasty.

Aims: The primary objective of this registry-based study was to compare patient-reported outcomes of cementless and cemented medial unicompartmental knee arthroplasty (UKA) during the first postoperative year. The secondary objective was to assess one- and three-year implant survival of both fixation techniques. Methods: We analyzed 10,862 cementless and 7,917 cemented UKA cases enrolled in the Dutch Arthroplasty Registry, operated between 2017 and 2021. Pre- to postoperative change in outcomes at six and 12 months' follow-up were compared using mixed model analyses. Kaplan-Meier and Cox regression models were applied to quantify differences in implant survival. Adjustments were made for patient-specific variables and annual hospital volume. Results: Change from baseline in the Oxford Knee Score (OKS) and activity-related pain was comparable between groups. Adjustment for covariates demonstrated a minimally greater decrease in rest-related pain in the cemented group (ß = -0.09 (95% confidence interval (CI) -0.16 to -0.01)). Cementless fixation was associated with a higher probability of achieving an excellent OKS outcome (> 41 points) (adjusted odds ratio 1.2 (95% CI 1.1 to 1.3)). The likelihood of one-year implant survival was greater for cemented implants (adjusted hazard ratio (HR) 1.35 (95% CI 1.01 to 1.71)), with higher revision rates for periprosthetic fractures of cementless implants. During two to three years' follow-up, the likelihood of implant survival was non-significantly greater for cementless UKA (adjusted HR 0.64 (95% CI 0.40 to 1.04)), primarily due to increased revision rates for tibial loosening of cemented implants. Conclusion: Cementless and cemented medial UKA led to comparable improvement in physical function and pain reduction during the initial postoperative year, albeit with a greater likelihood of achieving excellent OKS outcomes after cementless UKA. Anticipated differences in early physical function and pain should not be a decisive factor in the choice of fixation technique. However, surgeons should consider the differences in short- and long-term implant survival when deciding which implant to use.

Robotic-arm-assisted lateral unicompartmental knee arthroplasty leads to high implant survival and patient satisfaction at mean 10-year follow-up.

BACKGROUND: There is a lack of literature reporting on long-term outcomes following robotic-arm-assisted lateral unicompartmental knee arthroplasty (UKA). This study assessed the long-term survivorship, patient-reported satisfaction and pain scores following robotic-arm-assisted lateral UKA for lateral compartment osteoarthritis (OA). METHODS: A single surgeon's database was reviewed to identify all patients who underwent robotic-arm-assisted lateral UKA with a cemented, fixed-bearing prosthesis prior to May 2015. Patients were contacted to determine implant survivorship, satisfaction and pain. Kaplan-Meier models were applied to analyse survival. RESULTS: A total of 77 knees (70 patients) with a mean follow-up of 10.2 ± 1.5 years (range: 8.1-13.3) were included. Five knees were revised, corresponding to a 10-year survivorship of 96.1% and estimated survival time of 12.7 ± 0.3 years (95% confidence interval: 12.2-13.2) with all-cause revision as the endpoint. Unexplained pain (40.0%) and progression of OA (40.0%) in contralateral compartments were the most reported reasons for revision. Among patients without revision, 94.4% were either satisfied or very satisfied with their lateral UKA and the average pain score was 1.1. CONCLUSION: Robotic-arm-assisted lateral UKA led to high implant survivorship and patient satisfaction, and low pain scores at long-term follow-up. Progression of OA in contralateral compartments and unexplained pain were the most frequent reasons for revision. These findings support the continued use of robotic-arm-assisted lateral UKA for lateral compartment OA; however, its clinical value over conventional techniques remains to be established in prospective comparative studies. LEVEL OF EVIDENCE: Therapeutic Level IV.

Statistical shape analysis and computational modeling reveal novel relationships between tibiofemoral bony geometry and knee mechanics in young, female athletes.

Young female athletes participating in sports requiring rapid changes of direction are at heightened risk of suffering traumatic knee injury, especially noncontact rupture of the anterior cruciate ligament (ACL). Clinical studies have revealed that geometric features of the tibiofemoral joint are associated with increased risk of suffering noncontact ACL injury. However, the relationship between three-dimensional (3D) tibiofemoral geometry and knee mechanics in young female athletes is not well understood. We developed a statistically augmented computational modeling workflow to determine relationships between 3D geometry of the knee and tibiofemoral kinematics and ACL force in response to an applied loading sequence of compression, valgus, and anterior force, which is known to load the ACL. This workflow included 3D characterization of tibiofemoral bony geometry via principal component analysis and multibody dynamics models incorporating subject-specific knee geometries. A combination of geometric features of both the tibia and the femur that spanned all three anatomical planes was related to increased ACL force and to increased kinematic coupling (i.e., anterior, medial, and distal tibial translations and internal tibial rotation) in response to the applied loads. In contrast, a uniplanar measure of tibiofemoral geometry that is associated with ACL injury risk, sagittal plane slope of the lateral tibial plateau subchondral bone, was not related to ACL force. Thus, our workflow may aid in developing mechanics-based ACL injury screening tools for young, active females based on a unique combination of bony geometric features that are related to increased ACL loading.

A practical guide to the development and deployment of deep learning models for the orthopaedic surgeon: Part III, focus on registry creation, diagnosis, and data privacy.

Deep learning is a subset of artificial intelligence (AI) with enormous potential to transform orthopaedic surgery. As has already become evident with the deployment of Large Language Models (LLMs) like ChatGPT (OpenAI Inc.), deep learning can rapidly enter clinical and surgical practices. As such, it is imperative that orthopaedic surgeons acquire a deeper understanding of the technical terminology, capabilities and limitations associated with deep learning models. The focus of this series thus far has been providing surgeons with an overview of the steps needed to implement a deep learning-based pipeline, emphasizing some of the important technical details for surgeons to understand as they encounter, evaluate or lead deep learning projects. However, this series would be remiss without providing practical examples of how deep learning models have begun to be deployed and highlighting the areas where the authors feel deep learning may have the most profound potential. While computer vision applications of deep learning were the focus of Parts I and II, due to the enormous impact that natural language processing (NLP) has had in recent months, NLP-based deep learning models are also discussed in this final part of the series. In this review, three applications that the authors believe can be impacted the most by deep learning but with which many surgeons may not be familiar are discussed: (1) registry construction, (2) diagnostic AI and (3) data privacy. Deep learning-based registry construction will be essential for the development of more impactful clinical applications, with diagnostic AI being one of those applications likely to augment clinical decision-making in the near future. As the applications of deep learning continue to grow, the protection of patient information will become increasingly essential; as such, applications of deep learning to enhance data privacy are likely to become more important than ever before. Level of Evidence: Level IV.

Medical Imaging Applications Developed Using Artificial Intelligence Demonstrate High Internal Validity Yet Are Limited in Scope and Lack External Validation.

PURPOSE: To (1) review definitions and concepts necessary to interpret applications of deep learning (DL; a domain of artificial intelligence that leverages neural networks to make predictions on media inputs such as images) and (2) identify knowledge and translational gaps in the literature to provide insight into specific areas for improvement as adoption of this technology continues. METHODS: A comprehensive search of the literature was performed in December 2023 for articles regarding the use of DL in sports medicine. For each study, information regarding the joint of focus, specific anatomic structure/pathology to which DL was applied, imaging modality utilized, source of images used for model training and testing, data set size, model performance, and whether the DL model was externally validated was recorded. A numerical scale was used to rate each DL model's clinical impact, with 1 corresponding to proof-of-concept studies with little to no direct clinical impact and 5 corresponding to practice-changing clinical impact and ready for clinical deployment. RESULTS: Fifty-five studies were identified, all of which were published within the past 5 years, while 82% were published within the past 3 years. Of the DL models identified, 84% were developed for classification tasks, 9% for automated measurements, and 7% for segmentation. A total of 62% of studies utilized magnetic resonance imaging as the imaging modality, 25% radiographs, and 7% ultrasound, while 1 study each used computed tomography, arthroscopic images, or arthroscopic video. Sixty-five percent of studies focused on the detection of tears (anterior cruciate ligament [ACL], rotator cuff [RC], and meniscus). The diagnostic performance of ACL tears, as determined by the area under the receiver operator curve (AUROC), ranged from 0.81 to 0.99 for ACL tears (excellent to near perfect), 0.83 to 0.94 for RC tears (excellent), and from 0.75 to 0.96 for meniscus tears (acceptable to excellent). In addition, 3 studies focused on detection of cartilage lesions had AUROC ranging from 0.90 to 0.92 (excellent performance). However, only 4 (7%) studies externally validated their models, suggesting that they may not be generalizable or may not perform well when applied to populations other than that used to develop the model. Finally, the mean clinical impact score was 2 (range, 1-3) on scale of 1 to 5, corresponding to limited clinical applicability. CONCLUSIONS: DL models in orthopaedic sports medicine show generally excellent performance (high internal validity) but require external validation to facilitate clinical deployment. In addition, current models have low clinical applicability and fail to advance the field due to a focus on routine tasks and a narrow conceptual framework. LEVEL OF EVIDENCE: Level IV, scoping review of Level I to IV studies.

Independent data collectors decrease bias in the measurement of adherence to anterior cruciate ligament injury prevention programs.

OBJECTIVES: Studies on adherence to neuromuscular training (NMT) for anterior cruciate ligament (ACL) injury prevention are frequently biased due to the use of self-reporting by coaches or the athletes themselves. Few NMT studies use data collectors (aside from the athletes or the individuals administering the NMT program) to decrease bias when assessing the adherence of coaches and sports teams. We hypothesized that the use of a data collector who is independent of the team to evaluate adherence to NMT programs would be reliable. METHODS: In a prior a cluster-randomized controlled trial evaluating adherence to NMT training trial, twelve boys' and nine girls' high school athletic teams in a variety of sports were enrolled. Eight data collectors (unaffiliated with the NMT program) were hired specifically to record adherence of the athletes to the NMT exercises at each team's warm-ups 2-3 times a week, prior to practices and games. In addition to the data collectors, a control group of independent observers made visits throughout the season to also record adherence (solely for the purpose of this study, alongside the data collectors and in the same fashion) in order to evaluate the data collectors' performance and determine inter-observer reliability. The inter-observer reliability between data collectors and independent observers was measured using the Kappa statistic. RESULTS: A total of 399 warm-ups for practices or games were observed by data collectors to obtain adherence data. Independent observers also measured adherence at 58 practices or games for inter-observer reliability. Exercise instruction and alignment cues for 29 different exercises were analysed. The Kappa values ranged from 0.63 to 1.0, indicating substantial to perfect agreement. The overall Kappa values of 0.89 and 0.90 for exercise instruction and alignment cues, respectively, indicated almost perfect agreement. CONCLUSION: The use of a data collector who is independent of the team to evaluate adherence to NMT programs (rather than athlete or coach self-reporting), was shown to be a reliable method for measurement of adherence in studies of NMT for injury prevention. Avoiding self-reporting in adherence research to NMT training may decrease bias. LEVEL OF EVIDENCE: I.

The phenotypic diversity of anteromedial osteoarthritis before and after treatment with medial unicompartmental knee arthroplasty: A radiographic analysis of 1000 knees.

PURPOSE: This study aimed to assess phenotypic variation in the coronal plane of knees with anteromedial osteoarthritis using the functional knee phenotype classification, before and after treatment with medial unicompartmental knee arthroplasty (UKA). METHODS: The study comprised 1000 knees of 835 patients (45% females, 55% males, 90% Caucasian) who underwent medial UKA for anteromedial osteoarthritis. Pre and postoperative alignment was evaluated through the hip-knee-ankle angle (HKA), femoral mechanical angle (FMA), and tibial mechanical angle (TMA). Knees were classified according to the functional knee phenotype system which combines limb phenotype (HKA), and femoral and tibial knee phenotypes (FMA and TMA, respectively). Restoration of prearthritic coronal alignment following medial UKA was evaluated by phenotype. RESULTS: Preoperatively, 76 distinct and 25 relevant (prevalence ≥1%) functional knee phenotypes were identified, of which VARHKA 6°VARFMA 3°NEUTMA 0° was the most common (9.4% of knees). The most prevalent limb phenotype, VARHKA 6°, comprised 15 distinct knee phenotypes (FMA and TMA combinations). Postoperatively, 58 distinct and 17 relevant functional knee phenotypes were observed, of which VARHKA 3°NEUFMA 0°NEUTMA 0° had the highest prevalence at 18.3%. Knees with combined tibial and femoral deformities were associated with a lower probability of restoration of prearthritic coronal alignment following medial UKA, compared to knees without extra-articular deformity, or knees with an isolated tibial or femoral deformity. CONCLUSION: Phenotype analysis using the functional knee phenotype system demonstrated a wide diversity of coronal alignment phenotypes among knees with anteromedial osteoarthritis in a predominantly Caucasian population. Following medial UKA, a reduction from 25 preoperative to 17 postoperative relevant phenotypes was observed. Consideration of phenotypic variation can be of importance when aiming to restore prearthritic coronal alignment during medial UKA. LEVEL OF EVIDENCE: Level III, retrospective cohort study.

Role of Lateral Extra-articular Tenodesis in Restraining Internal Tibial Rotation: In Vitro Biomechanical Assessment of Lateral Tissue Engagement.

BACKGROUND: The way in which force increases in the anterolateral tissues and the lateral extra-articular tenodesis (LET) tissue to resist internal rotation (IR) of the tibia after anterior cruciate ligament (ACL) reconstruction in isolation and after LET augmentation, respectively, is not well understood. PURPOSE: (1) To compare in a cadaveric model how force increases (ie, engages) in the anterolateral tissues with IR of the tibia after isolated ACL reconstruction and in the LET tissue after augmentation of the ACL reconstruction with LET and (2) to determine whether IR of the tibia is related to engagement of the LET tissue. STUDY DESIGN: Controlled laboratory study. METHODS: IR moments were applied to 9 human cadaveric knees at 0°, 30°, 60°, and 90° of flexion using a robotic manipulator. Each knee was tested in 2 states: (1) after isolated ACL reconstruction with intact anterolateral tissues and (2) after LET was performed using a modified Lemaire technique with the LET tissue fixed at 60° of flexion under 44 N of tension. Resultant forces carried by the anterolateral tissues and the LET tissue were determined via superposition. The way force increased in these tissues was characterized via parameters of tissue engagement, namely in situ slack, in situ stiffness, and tissue force at peak applied IR moment, and then compared (α < .05). IR was related to parameters of engagement of the LET tissue via simple linear regression (α < .05). RESULTS: The LET tissue exhibited less in situ slack than the anterolateral tissues at 30°, 60°, and 90° of flexion (P≤ .04) and greater in situ stiffness at 30° and 90° of flexion (P≤ .043). The LET tissue carried greater force at the peak applied IR moment at 0° and 30° of flexion (P≤ .01). IR was related to the in situ slack of the LET tissue (R2≥ 0.88; P≤ .0003). CONCLUSION: LET increased restraint to IR of the tibia compared with the anterolateral tissue, particularly at 30°, 60°, and 90° of flexion. IR of the tibia was positively associated with in situ slack of the LET tissue. CLINICAL RELEVANCE: Fixing the LET at 60° of flexion still provided IR restraint in the more functionally relevant flexion angle of 30°. Surgeons should pay close attention to the angle of internal and/or external tibial rotation when fixing the LET tissue intraoperatively because this surgical parameter is related to in situ slack of the LET tissue and, therefore, the amount of IR of the tibia.

De Novo Synovial Chondromatosis following Primary Total Knee Arthroplasty: A Case Report.

In this case report, we present a rare case of a female patient who developed pain and swelling after a total knee arthroplasty. An extensive diagnostic workup including serum and synovial testing to rule out infection was performed in addition to advanced imaging including an MRI of the knee, but it was only after an arthroscopic synovectomy that the diagnosis of secondary synovial chondromatosis was confirmed. The purpose of this case report is to highlight the occurrence of secondary synovial chondromatosis as a rare cause of pain and swelling after total knee arthroplasty, thereby assisting clinicians in providing prompt diagnosis, surgical treatment, and efficient recovery in the setting of secondary synovial chondromatosis after total knee arthroplasty.

Restoration or relative overcorrection of pre-arthritic coronal alignment leads to improved results following medial unicompartmental knee arthroplasty.

PURPOSE: A pre-arthritic alignment strategy for medial unicompartmental knee arthroplasty (UKA) aims to restore a patient's native lower limb alignment which may translate into improved outcomes. This study aimed to assess whether patients with pre-arthritically aligned knees versus patients with non-pre-arthritically aligned knees demonstrated improved mid-term outcomes and survivorship following medial UKA. The hypothesis was that pre-arthritic alignment in medial UKA would lead to better postoperative outcomes. METHODS: A retrospective study of 537 robotic-assisted fixed-bearing medial UKA was conducted. During this procedure, the surgical goal was to restore pre-arthritic alignment guided by re-tensioning of the medial collateral ligament (MCL). For study purposes, coronal alignment was retrospectively evaluated using the mechanical hip-knee-ankle angle (mHKA). Pre-arthritic alignment was estimated through the arithmetic hip-knee-ankle (aHKA) algorithm. Knees were grouped according to the difference between postoperative mHKA and estimated pre-arthritic alignment (i.e., mHKA - aHKA) as Group 1 (pre-arthritically aligned: mHKA restored within 2.0° of the aHKA), Group 2 (mHKA > 2.0° overcorrected relative to the aHKA), or Group 3 (mHKA > 2.0° undercorrected relative to the aHKA). Outcomes included the Knee Injury and Osteoarthritic Outcome Score for Joint Replacement (KOOS, JR), Kujala, proportions of knees achieving the patient acceptable symptom state (PASS) for these scores, and survivorship. PASS thresholds for KOOS, JR and Kujala were determined using a receiver operating characteristic curve method. RESULTS: A total of 369 knees were categorized as Group 1, 107 as Group 2, and 61 as Group 3. At 4.4 ± 1.6 years follow-up, mean KOOS, JR was comparable among groups, while Kujala was significantly worse in Group 3. The proportion of knees achieving the PASS for Kujala (76.5 points) was lower in Group 3 (n = 32; 59%) compared to Group 1 (n = 260; 74%) (p = 0.02). 5-year survivorship was higher in Group 1 and Group 2 (99% and 100%, respectively) compared to Group 3 (91%) (p = 0.04). CONCLUSION: Pre-arthritically aligned knees and knees with relative overcorrection from their pre-arthritic alignment following medial UKA demonstrated improved mid-term outcomes and survivorship compared to knees with relative under correction from their pre-arthritic alignment. These results encourage restoring or relatively overcorrecting pre-arthritic alignment to optimize outcomes following medial UKA, and caution against under correction from the pre-arthritic alignment. LEVEL OF EVIDENCE: IV, case series.

Ten-Year Survivorship and Patient Satisfaction Following Robotic-Arm-Assisted Medial Unicompartmental Knee Arthroplasty: A Prospective Multicenter Study.

BACKGROUND: Robotic-arm-assisted unicompartmental knee arthroplasty (UKA) has been shown to result in high short- and mid-term survivorship. However, it is not known whether these outcomes are maintained at long-term follow-up. This study aimed to evaluate long-term implant survivorship, modes of failure, and patient satisfaction following robotic-arm-assisted medial UKA. METHODS: A prospective multicenter study of 474 consecutive patients (531 knees) undergoing robotic-arm-assisted medial UKA was conducted. A cemented, fixed-bearing system with a metal-backed onlay tibial implant was used in all cases. Patients were contacted at 10-year follow-up to determine implant survivorship and satisfaction. Survival was analyzed using Kaplan-Meier models. RESULTS: Data were analyzed for 366 patients (411 knees) with a mean follow-up of 10.2 ± 0.4 years. A total of 29 revisions were reported, corresponding to a 10-year survivorship of 91.7% (95% confidence interval, 88.8% to 94.6%). Of all revisions, 26 UKAs were revised to total knee arthroplasty. Unexplained pain and aseptic loosening were the most commonly reported modes of failure, accounting for 38% and 35% of revisions, respectively. Of patients without revision, 91% were either satisfied or very satisfied with their overall knee function. CONCLUSIONS: This prospective multicenter study found high 10-year survivorship and patient satisfaction following robotic-arm-assisted medial UKA. Pain and fixation failure remained common causes for revision following cemented fixed-bearing medial UKA, despite the use of a robotic-arm-assisted technique. Prospective comparative studies are needed to assess the clinical value of robotic assistance over conventional techniques in UKA. LEVEL OF EVIDENCE: Prognostic Level II . See Instructions for Authors for a complete description of levels of evidence.

Key Thresholds and Relative Contributions of Knee Geometry, Anteroposterior Laxity, and Body Weight as Risk Factors for Noncontact ACL Injury.

Background: Limited data exist regarding the association of tibiofemoral bony and soft tissue geometry and knee laxity with risk of first-time noncontact anterior cruciate ligament (ACL) rupture. Purpose: To determine associations of tibiofemoral geometry and anteroposterior (AP) knee laxity with risk of first-time noncontact ACL injury in high school and collegiate athletes. Study Design: Cohort study; Level of evidence, 2. Methods: Over a 4-year period, noncontact ACL injury events were identified as they occurred in 86 high school and collegiate athletes (59 female, 27 male). Sex- and age-matched control participants were selected from the same team. AP laxity of the uninjured knee was measured using a KT-2000 arthrometer. Magnetic resonance imaging was taken on ipsilateral and contralateral knees, and articular geometries were measured. Sex-specific general additive models were implemented to investigate associations between injury risk and 6 features: ACL volume, meniscus-bone wedge angle in the lateral compartment of the tibia, articular cartilage slope at the middle region of the lateral compartment of the tibia, femoral notch width at the anterior outlet, body weight, and AP displacement of the tibia relative to the femur. Importance scores (in percentages) were calculated to rank the relative contribution of each variable. Results: In the female cohort, the 2 features with the highest importance scores were tibial cartilage slope (8.6%) and notch width (8.1%). In the male cohort, the 2 top-ranked features were AP laxity (5.6%) and tibial cartilage slope (4.8%). In female patients, injury risk increased by 25.5% with lateral middle cartilage slope becoming more posteroinferior from -6.2° to -2.0° and by 17.5% with lateral meniscus-bone wedge angle increasing from 27.3° to 28.2°. In males, an increase in AP displacement from 12.5 to 14.4 mm in response to a 133-N anterior-directed load was associated with a 16.7% increase in risk. Conclusion: Of the 6 variables studied, there was no single dominant geometric or laxity risk factor for ACL injury in either the female or male cohort. In males, AP laxity >13 to 14 mm was associated with sharply increased risk of noncontact ACL injury. In females, lateral meniscus-bone wedge angle >28° was associated with a sharply decreased risk of noncontact ACL injury.

Higher Adherence to Anterior Cruciate Ligament Injury Prevention Programs Is Associated With Lower Injury Rates: A Meta-Analysis and Meta-Regression.

Background: Athletes who participate in sports that involve cutting and pivoting movements are particularly susceptible to anterior cruciate ligament (ACL) injury. Preventing this injury is the best way to combat its health consequences and costs. There may be a dose-response relationship between adherence and injury reduction. Purpose: We sought to examine whether athletes' adherence to injury prevention programs (IPPs) is associated with reductions in ACL and lower extremity (LE) injuries. Methods: We conducted a systematic review of the PubMed, EMBASE, and Cochrane Library databases, searching for studies published between 2011 and 2021. Studies were included if they reported on the use of an ACL IPP compared with a control group and recorded the rate of injuries to calculate a rate ratio, as well as adherence to the program as a percentage of sessions performed. For the meta-analysis, the rate ratios were pooled using the DerSimonian-Laird random-effects model. Results: For the 15 studies included (11 randomized controlled trials and 4 cohort studies), the random-effects model grouped athletes' adherence to an IPP as high (76% or more of the sessions), moderate (51%-75% of the sessions), and low (50% or fewer of the sessions). We found that athletes with the highest level of IPP adherence had a significantly lower incidence of ACL injury. The rate ratios for moderate and low adherence did not demonstrate a reduced incidence of ACL injury. Injury prevention program participation was also associated with a decrease in LE injury rates. Conclusion: This systematic review and meta-analysis found that athletes with high adherence to IPPs had reduced rates of ACL and LE injuries. Our findings suggest that educating coaches and athletes on the dose-dependent benefits of IPPs may promote the routine incorporation of these programs into warm-up sessions to decrease the risk of ACL and LE injuries.

Survival analyses and their applications in orthopaedics.

Survival analyses are a powerful statistical tool used to analyse data when the outcome of interest involves the time until an event. There is an array of models fit for this goal; however, there are subtle differences in assumptions, as well as a number of pitfalls, that can lead to biased results if researchers are unaware of the subtleties. As larger amounts of data become available, and more survival analyses are published every year, it is important that healthcare professionals understand how to evaluate these models and apply them into their practice. Therefore, the purpose of this study was to present an overview of survival analyses, including required assumptions and important pitfalls, as well as examples of their use within orthopaedic surgery.

Artificial Intelligence for Automated Implant Identification in Knee Arthroplasty: A Multicenter External Validation Study Exceeding 3.5 Million Plain Radiographs.

BACKGROUND: Surgical management of complications following knee arthroplasty demands accurate and timely identification of implant manufacturer and model. Automated image processing using deep machine learning has been previously developed and internally validated; however, external validation is essential prior to scaling clinical implementation for generalizability. METHODS: We trained, validated, and externally tested a deep learning system to classify knee arthroplasty systems as one of the 9 models from 4 manufacturers derived from 4,724 original, retrospectively collected anteroposterior plain knee radiographs across 3 academic referral centers. From these radiographs, 3,568 were used for training, 412 for validation, and 744 for external testing. Augmentation was applied to the training set (n = 3,568,000) to increase model robustness. Performance was determined by the area under the receiver operating characteristic curve, sensitivity, specificity, and accuracy. Implant identification processing speed was calculated. The training and testing sets were drawn from statistically different populations of implants (P < .001). RESULTS: After 1,000 training epochs by the deep learning system, the system discriminated 9 implant models with a mean area under the receiver operating characteristic curve of 0.989, accuracy of 97.4%, sensitivity of 89.2%, and specificity of 99.0% in the external testing dataset of 744 anteroposterior radiographs. The software classified implants at a mean speed of 0.02 seconds per image. CONCLUSION: An artificial intelligence-based software for identifying knee arthroplasty implants demonstrated excellent internal and external validation. Although continued surveillance is necessary with implant library expansion, this software represents a responsible and meaningful clinical application of artificial intelligence with immediate potential to globally scale and assist in preoperative planning prior to revision knee arthroplasty.

A practical guide to the development and deployment of deep learning models for the orthopedic surgeon: part II.

Deep learning has the potential to be one of the most transformative technologies to impact orthopedic surgery. Substantial innovation in this area has occurred over the past 5 years, but clinically meaningful advancements remain limited by a disconnect between clinical and technical experts. That is, it is likely that few orthopedic surgeons possess both the clinical knowledge necessary to identify orthopedic problems, and the technical knowledge needed to implement deep learning-based solutions. To maximize the utilization of rapidly advancing technologies derived from deep learning models, orthopedic surgeons should understand the steps needed to design, organize, implement, and evaluate a deep learning project and its workflow. Equipping surgeons with this knowledge is the objective of this three-part editorial review. Part I described the processes involved in defining the problem, team building, data acquisition, curation, labeling, and establishing the ground truth. Building on that, this review (Part II) provides guidance on pre-processing and augmenting the data, making use of open-source libraries/toolkits, and selecting the required hardware to implement the pipeline. Special considerations regarding model training and evaluation unique to deep learning models relative to "shallow" machine learning models are also reviewed. Finally, guidance pertaining to the clinical deployment of deep learning models in the real world is provided. As in Part I, the focus is on applications of deep learning for computer vision and imaging.

Does a Uni "Feel Better" than a Total Knee? Not Necessarily, When Using Modern Implant Designs.

Background: When comparing functional outcomes of patients with unicompartmental knee arthroplasty (UKA) versus total knee arthroplasty (TKA), studies often report the UKA as the preferred procedure; however, recent improvements in the design of modern TKA implants have aimed at narrowing this gap. Purpose: We sought to compare the "feel" of modern TKA implants to that of UKA, using the Forgotten Joint Score (FJS), a validated patient-reported outcome measure. Methods: We performed a retrospective review of patients who underwent TKA and UKA at 2 institutions between 2014 and 2017. All UKA procedures were robotic arm-assisted with a single implant, "traditional TKAs" were performed using traditional posterior-stabilized implants, and "modern TKAs" were performed using posterior-stabilized implants with a modern design. Differences in FJS were assessed using 1-way analysis of variance and independent 2-sample t tests. Results: A total of 600 patients were included in our study, with 200 patients in each surgical subcategory. Mean age was 62.8 ± 10.2 years and mean body mass index was 29.9 ± 4.9. Modern TKA and UKA had similar FJS at 1 year. While modern TKA had a significantly higher FJS than traditional TKA, UKA did not have a significantly higher FJS than traditional TKA. Conclusion: Our retrospective analysis found no significant differences in the FJS of patients who underwent UKA and TKA with a modern design; however, both had superior scores than traditional TKA designs. This finding suggests that modern TKA designs may have the potential to achieve the natural feeling that is typically associated with joint-conserving surgeries such as UKA, although longer follow-up is necessary.

Bayesian Calibration of Computational Knee Models to Estimate Subject-Specific Ligament Properties, Tibiofemoral Kinematics, and Anterior Cruciate Ligament Force With Uncertainty Quantification.

High-grade knee laxity is associated with early anterior cruciate ligament (ACL) graft failure, poor function, and compromised clinical outcome. Yet, the specific ligaments and ligament properties driving knee laxity remain poorly understood. We described a Bayesian calibration methodology for predicting unknown ligament properties in a computational knee model. Then, we applied the method to estimate unknown ligament properties with uncertainty bounds using tibiofemoral kinematics and ACL force measurements from two cadaver knees that spanned a range of laxities; these knees were tested using a robotic manipulator. The unknown ligament properties were from the Bayesian set of plausible ligament properties, as specified by their posterior distribution. Finally, we developed a calibrated predictor of tibiofemoral kinematics and ACL force with their own uncertainty bounds. The calibrated predictor was developed by first collecting the posterior draws of the kinematics and ACL force that are induced by the posterior draws of the ligament properties and model parameters. Bayesian calibration identified unique ligament slack lengths for the two knee models and produced ACL force and kinematic predictions that were closer to the corresponding in vitro measurement than those from a standard optimization technique. This Bayesian framework quantifies uncertainty in both ligament properties and model outputs; an important step towards developing subject-specific computational models to improve treatment for ACL injury.

Robotic-assisted medial unicompartmental knee arthroplasty restores estimated pre-arthritic coronal limb alignment: A retrospective cohort study.

BACKGROUND: Robotic-assisted medial unicompartmental knee arthroplasty (UKA) aims to restore pre-arthritic (constitutional) limb alignment, by re-tensioning of the medial collateral ligament (MCL). This study aimed to determine whether pre-arthritic coronal alignment was restored following robotic-assisted medial UKA in patients with medial compartment osteoarthritis. METHOD: A retrospective study was undertaken, including 102 patients with a unilateral robotic-assisted medial UKA and a contralateral unaffected knee. Both the validated arithmetic hip-knee-ankle angle (aHKA) and alignment of the contralateral unaffected knee were used to estimate pre-arthritic alignment. The aHKA is a radiographic method to estimate the pre-arthritic mechanical hip-knee-ankle angle (mHKA). To verify restoration of pre-arthritic alignment, postoperative mHKA was compared to the aHKA. Additionally, postoperative mHKA, joint line congruence (JLCA), and knee joint line obliquity (KJLO) angles were compared between the operative and contralateral unaffected knee. Equivalence between postoperative and pre-arthritic alignment was assessed through the two-one-sided t-test (TOST), using equivalence margins of ±2.0°. RESULTS: Postoperative mHKA was equivalent to the aHKA (mean difference -0.38°, 90% CI -0.69 to -0.07;p < .001), with 93 knees (91%) restored within 3.0° of their aHKA. Postoperative mHKA, JLCA and KJLO were equivalent between the operative and contralateral unaffected knees, with mean differences of -0.65°, -0.65°, and -0.40°, respectively; all p < .001. CONCLUSIONS: Postoperative and pre-arthritic coronal alignment were equivalent following robotic-assisted medial UKA, with 91% of knees restored within 3.0° of their pre-arthritic mechanical axis. These results demonstrate that both mechanical alignment and joint line congruence are restored by MCL re-tensioning in patients undergoing robotic-assisted medial UKA for medial compartment osteoarthritis.