Aims

Machine learning (ML), a branch of artificial intelligence that uses algorithms to learn from data and make predictions, offers a pathway towards more personalized and tailored surgical treatments. This approach is particularly relevant to prevalent joint diseases such as osteoarthritis (OA). In contrast to end-stage disease, where joint arthroplasty provides excellent results, early stages of OA currently lack effective therapies to halt or reverse progression. Accurate prediction of OA progression is crucial if timely interventions are to be developed, to enhance patient care and optimize the design of clinical trials.

The June 2024 Wrist & Hand Roundup³⁶⁰ looks at: One-year outcomes of the anatomical front and back reconstruction for scapholunate dissociation; Limited intercarpal fusion versus proximal row carpectomy in the treatment of SLAC or SNAC wrist: results after 3.5 years; Prognostic factors for clinical outcomes after arthroscopic treatment of traumatic central tears of the triangular fibrocartilage complex; The rate of nonunion in the MRI-detected occult scaphoid fracture: a multicentre cohort study; Does correction of carpal malalignment influence the union rate of scaphoid nonunion surgery?; Provision of a home-based video-assisted therapy programme in thumb carpometacarpal arthroplasty; Is replantation associated with better hand function after traumatic hand amputation than after revision amputation?; Diagnostic performance of artificial intelligence for detection of scaphoid and distal radius fractures: a systematic review.

Aims. This study was designed to develop a model for predicting bone mineral density (BMD) loss of the femur after total hip arthroplasty (THA) using artificial intelligence (AI), and to identify factors that influence the prediction. Additionally, we virtually examined the efficacy of administration of bisphosphonate for cases with severe BMD loss based on the predictive model. Methods. The study included 538 joints that underwent primary THA. The patients were divided into groups using unsupervised time series clustering for five-year BMD loss of Gruen zone 7 postoperatively, and a machine-learning model to predict the BMD loss was developed. Additionally, the predictor for BMD loss was extracted using SHapley Additive exPlanations (SHAP). The patient-specific efficacy of bisphosphonate, which is the most important categorical predictor for BMD loss, was examined by calculating the change in predictive probability when hypothetically switching between the inclusion and exclusion of bisphosphonate. Results. Time series clustering allowed us to divide the patients into two groups, and the predictive factors were identified including patient- and operation-related factors. The area under the receiver operating characteristic (ROC) curve (AUC) for the BMD loss prediction averaged 0.734. Virtual administration of bisphosphonate showed on average 14% efficacy in preventing BMD loss of zone 7. Additionally, stem types and preoperative triglyceride (TG), creatinine (Cr), estimated glomerular filtration rate (eGFR), and creatine kinase (CK) showed significant association with the estimated patient-specific efficacy of bisphosphonate. Conclusion. Periprosthetic BMD loss after THA is predictable based on patient- and operation-related factors, and optimal prescription of bisphosphonate based on the prediction may prevent BMD loss. Cite this article: Bone Joint Res 2024;13(4):184–192

Anterior approach total hip arthroplasty (AA-THA) has a steep learning curve, with higher complication rates in initial cases. Proper surgical case selection during the learning curve can reduce early risk. This study aims to identify patient and radiographic factors associated with AA-THA difficulty using Machine Learning (ML). Consecutive primary AA-THA patients from two centres, operated by two expert surgeons, were enrolled (excluding patients with prior hip surgery and first 100 cases per surgeon). K- means prototype clustering – an unsupervised ML algorithm – was used with two variables - operative duration and surgical complications within 6 weeks - to cluster operations into difficult or standard groups. Radiographic measurements (neck shaft angle, offset, LCEA, inter-teardrop distance, Tonnis grade) were measured by two independent observers. These factors, alongside patient factors (BMI, age, sex, laterality) were employed in a multivariate logistic regression analysis and used for k-means clustering. Significant continuous variables were investigated for predictive accuracy using Receiver Operator Characteristics (ROC). Out of 328 THAs analyzed, 130 (40%) were classified as difficult and 198 (60%) as standard. Difficult group had a mean operative time of 106mins (range 99–116) with 2 complications, while standard group had a mean operative time of 77mins (range 69–86) with 0 complications. Decreasing inter-teardrop distance (odds ratio [OR] 0.97, 95% confidence interval [CI] 0.95–0.99, p = 0.03) and right-sided operations (OR 1.73, 95% CI 1.10–2.72, p = 0.02) were associated with operative difficulty. However, ROC analysis showed poor predictive accuracy for these factors alone, with area under the curve of 0.56. Inter-observer reliability was reported as excellent (ICC >0.7). Right-sided hips (for right-hand dominant surgeons) and decreasing inter-teardrop distance were associated with case difficulty in AA-THA. These data could guide case selection during the learning phase. A larger dataset with more complications may reveal further factors

Aims. Knee osteoarthritis (OA) involves a variety of tissues in the joint. Gene expression profiles in different tissues are of great importance in order to understand OA. Methods. First, we obtained gene expression profiles of cartilage, synovium, subchondral bone, and meniscus from the Gene Expression Omnibus (GEO). Several datasets were standardized by merging and removing batch effects. Then, we used unsupervised clustering to divide OA into three subtypes. The gene ontology and pathway enrichment of three subtypes were analyzed. CIBERSORT was used to evaluate the infiltration of immune cells in different subtypes. Finally, OA-related genes were obtained from the Molecular Signatures Database for validation, and diagnostic markers were screened according to clinical characteristics. Quantitative reverse transcription polymerase chain reaction (qRT‐PCR) was used to verify the effectiveness of markers. Results. C1 subtype is mainly concentrated in the development of skeletal muscle organs, C2 lies in metabolic process and immune response, and C3 in pyroptosis and cell death process. Therefore, we divided OA into three subtypes: bone remodelling subtype (C1), immune metabolism subtype (C2), and cartilage degradation subtype (C3). The number of macrophage M0 and activated mast cells of C2 subtype was significantly higher than those of the other two subtypes. COL2A1 has significant differences in different subtypes. The expression of COL2A1 is related to age, and trafficking protein particle complex subunit 2 is related to the sex of OA patients. Conclusion. This study linked different tissues with gene expression profiles, revealing different molecular subtypes of patients with knee OA. The relationship between clinical characteristics and OA-related genes was also studied, which provides a new concept for the diagnosis and treatment of OA. Cite this article: Bone Joint Res 2023;12(12):702–711

Cite this article: Bone Joint Res 2023;12(8):494–496.

The use of artificial intelligence (AI) is rapidly growing across many domains, of which the medical field is no exception. AI is an umbrella term defining the practical application of algorithms to generate useful output, without the need of human cognition. Owing to the expanding volume of patient information collected, known as ‘big data’, AI is showing promise as a useful tool in healthcare research and across all aspects of patient care pathways. Practical applications in orthopaedic surgery include: diagnostics, such as fracture recognition and tumour detection; predictive models of clinical and patient-reported outcome measures, such as calculating mortality rates and length of hospital stay; and real-time rehabilitation monitoring and surgical training. However, clinicians should remain cognizant of AI’s limitations, as the development of robust reporting and validation frameworks is of paramount importance to prevent avoidable errors and biases. The aim of this review article is to provide a comprehensive understanding of AI and its subfields, as well as to delineate its existing clinical applications in trauma and orthopaedic surgery. Furthermore, this narrative review expands upon the limitations of AI and future direction.

Cite this article: Bone Joint Res 2023;12(7):447–454.

Primary total hip replacement (THR) is a successful and common operation which orthopaedic trainees must demonstrate competence in prior to completion of training. This study aimed to determine the impact of operating surgeon grade and level of supervision on the incidence of 1-year patient mortality and all-cause revision following elective primary THR in a large UK training centre. National Joint Registry (NJR) data for all elective primary THR performed in a single University Teaching Hospital from 2005–2020 were used, with analysis performed on the 15-year dataset divided into 5-year temporal periods (B1 2005–2010, B2 2010–2015, B3 2015–2020). Outcome measures were mortality and revision surgery at one year, in relation to lead surgeon grade, and level of supervision for trainee-led operations. 9999 eligible primary THR were undertaken, of which 5526 (55.3%) were consultant led (CL), and 4473 (44.7%) trainees led (TL). Of TL, 2404 (53.7%) were non-consultant supervised (TU), and 2069 (46.3%) consultant supervised (TS). The incidence of 1-year patient mortality was 2.05% (n=205), and all-cause revision was 1.11% (n=111). There was no difference in 1-year mortality between TL (n=82, 1.8%) and CL (n=123, 2.2%) operations (p=0.20, OR 0.78, CI 0.55–1.10). The incidence of 1-year revision was not different for TL (n=56, 1.3%) and CL (n=55, 1.0%) operations (p=0.15, OR 1.37, CI 0.89–2.09). Overall, there was no temporal change for either outcome measure between TL or CL operations. A significant increase in revision within 1-year was observed in B3 between TU (n=17, 2.7%) compared to CL (n=17, 1.0%) operations (p=0.005, OR 2.81, CI 1.35–5.87). We found no difference in 1-year mortality or 1-year all-cause revision rate between trainee-led primary THR and consultant-led operations over the entire fifteen-year period. However, unsupervised trainee led THR in the most recent 5-year block (2015–2020) has a significantly increased risk of early revision, mainly due to instability and prosthetic joint infection. This suggests that modern surgical training is having a detrimental effect on THR patient outcomes. More research is needed to understand the reasons if this trend is to be reversed

Aims

The aim of this study was to determine the effectiveness of home-based prehabilitation on pre- and postoperative outcomes in participants awaiting total knee (TKA) and hip arthroplasty (THA).

Depletion of Scleraxis-lineage (ScxLin) cells in adult tendon recapitulates age-related decrements in cell density, ECM organization and composition. However, depletion of ScxLin cells improves tendon healing, relative to age-matched wildtype mice, while aging impairs healing. Therefore, we examined whether ScxLin depletion and aging result in comparable shifts in the tendon cell environment and defined the intrinsic programmatic shifts that occur with natural aging, to define the key regulators of age-related healing deficits. ScxLin cells were depleted in 3M-old Scx-Cre+; Rosa-DTRF/+ mice via diphtheria toxin injections into the hindpaw. Rosa-DTRF/+ mice were used as wildtype (WT) controls. Tendons were harvested from 6M-old ScxLin depleted and WT mice, and 21-month-old (21M) C57Bl/6 mice (aged). FDL tendons (n=6) were harvested for single-cell RNAseq, pooled, collagenase digested, and sorted for single cell capture. Data was processed using Cell Ranger and then aligned to the annotated mouse genome (mm10). Filtering, unsupervised cell clustering, and differential gene expression (DEG) analysis were performed using Seurat. Following integration and sub-clustering of the tenocyte populations, five distinct subpopulations were observed. In both ScxLin depletion and aging, ‘ECM synthesizers’ and ‘ECM organizers’ populations were lost, consistent with disruptions in tissue homeostasis and altered ECM composition. However, in ScxLin depleted mice retention of a ‘specialized ECM remodeler’ population was observed, while aging tendon cells demonstrated inflammatory skewing with retention of a ‘pro-inflammatory tenocyte population’. In addition, enrichment of genes associated with protein misfolding clearance were observed in aged tenocytes. Finally, a similar inflammatory skewing was observed in aged tendon-resident macrophages, with this skewing not observed in ScxLin depleted tendons. These data suggest that loss of ‘ECM synthesizer’ populations underpins disruptions in tendon homeostasis. However, retention of ‘specialized remodelers’ promotes enhanced healing (ScxLin depletion), while inflammatory skewing may drive the impaired healing response in aged tendons

Physiotherapy is a critical element in successful conservative management of low back pain (LBP). The aim of this study was to develop and evaluate a system with wearable inertial sensors to objectively detect sitting postures and performance of unsupervised exercises containing movement in multiple planes (flexion, extension, rotation). A set of 8 inertial sensors were placed on 19 healthy adult subjects. Data was acquired as they performed 7 McKenzie low-back exercises and 3 sitting posture positions. This data was used to train two models (Random Forest (RF) and XGBoost (XGB)) using engineered time series features. In addition, a convolutional neural network (CNN) was trained directly on the time series data. A feature importance analysis was performed to identify sensor locations and channels that contributed most to the models. Finally, a subset of sensor locations and channels was included in a hyperparameter grid search to identify the optimal sensor configuration and the best performing algorithm(s) for exercise classification. Models were evaluated using F1-score in a 10-fold cross validation approach. The optimal hardware configuration was identified as a 3-sensor setup using lower back, left thigh, and right ankle sensors with acceleration, gyroscope, and magnetometer channels. The XBG model achieved the highest exercise (F1=0.94±0.03) and posture (F1=0.90±0.11) classification scores. The CNN achieved similar results with the same sensor locations, using only the accelerometer and gyroscope channels for exercise classification (F1=0.94±0.02) and the accelerometer channel alone for posture classification (F1=0.91±0.03). This study demonstrates the potential of a 3-sensor lower body wearable solution (e.g. smart pants) that can identify proper sitting postures and exercises in multiple planes, suitable for low back pain. This technology has the potential to improve the effectiveness of LBP rehabilitation by facilitating quantitative feedback, early problem diagnosis, and possible remote monitoring

Access to health care, including physiotherapy, is increasingly occurring through virtual formats. At-home adherence to physical therapy programs is often poor and few tools exist to objectively measure low back physiotherapy exercise participation without the direct supervision of a medical professional. The aim of this study was to develop and evaluate the potential for performing automatic, unsupervised video-based monitoring of at-home low back physiotherapy exercises using a single mobile phone camera. 24 healthy adult subjects performed seven exercises based on the McKenzie low back physiotherapy program while being filmed with two smartphone cameras. Joint locations were automatically extracted using an open-source pose estimation framework. Engineered features were extracted from the joint location time series and used to train a support vector machine classifier (SVC). A convolutional neural network (CNN) was trained directly on the joint location time series data to classify exercises based on a recording from a single camera. The models were evaluated using a 5-fold cross validation approach, stratified by subject, with the class-balanced accuracy used as the performance metric. Optimal performance was achieved when using a total of 12 pose estimation landmarks from the upper and lower body, with the SVC model achieving a classification accuracy of 96±4% and the CNN model an accuracy of 97±2%. This study demonstrates the feasibility of using a smartphone camera and a supervised machine learning model to effectively assess at-home low back physiotherapy adherence. This approach could provide a low-cost, scalable method for tracking adherence to physical therapy exercise programs in a variety of settings

Aims

An objective technological solution for tracking adherence to at-home shoulder physiotherapy is important for improving patient engagement and rehabilitation outcomes, but remains a significant challenge. The aim of this research was to evaluate performance of machine-learning (ML) methodologies for detecting and classifying inertial data collected during in-clinic and at-home shoulder physiotherapy exercise.

The February 2023 Knee Roundup. 360. looks at: Machine-learning models: are all complications predictable?; Positive cultures can be safely ignored in revision arthroplasty patients that do not meet the 2018 International Consensus Meeting Criteria; Spinal versus general anaesthesia in contemporary primary total knee arthroplasty; Preoperative pain and early arthritis are associated with poor outcomes in total knee arthroplasty; Risk factors for infection and revision surgery following patellar tendon and quadriceps tendon repairs; Supervised versus unsupervised rehabilitation following total knee arthroplasty; Kinematic alignment has similar outcomes to mechanical alignment: a systematic review and meta-analysis; Lifetime risk of revision after knee arthroplasty influenced by age, sex, and indication; Risk factors for knee osteoarthritis after traumatic knee injury

Giant cell tumors of bone (GCTs) are locally aggressive tumors with recurrence potential that represent up to 10% of primary tumors of the bone. GCTs pathogenesis is driven by neoplastic mononuclear stromal cells that overexpress receptor activator of nuclear factor kappa-B/ligand (RANKL). Treatment with specific anti-RANKL antibody (denosumab) was recently introduced, used either as a neo-adjuvant in resectable tumors or as a stand-alone treatment in unresectable tumors. While denosumab has been increasingly used, a percentage of patients do not improve after treatment. Here, we aim to determine molecular and histological patterns that would help predicting GCTs response to denosumab to improve personalized treatment. Nine pre-treatment biopsies of patients with spinal GCT were collected at 2 centres. In 4 patients denosumab was used as a neo-adjuvant, 3 as a stand-alone and 2 received denosumab as adjuvant treatment. Clinical data was extracted retrospectively. Total mRNA was extracted by using a formalin-fixed paraffin-embedded extraction kit and we determined the transcript profile of 730 immune-oncology related genes by using the Pan Cancer Immune Profiling panel (Nanostring). The gene expression was compared between patients with good and poor response to Denosumab treatment by using the nSolver Analysis Software (Nanostring). Immunohistochemistry was performed in the tissue slides to characterize cell populations and immune response in CGTs. Two out of 9 patients showed poor clinical response with tumor progression and metastasis. Our analysis using unsupervised hierarchical clustering determined differences in gene expression between poor responders and good responders before denosumab treatment. Poor responding lesions are characterized by increased expression of inflammatory cytokines as IL8, IL1, interferon a and g, among a myriad of cytokines and chemokines (CCL25, IL5, IL26, IL25, IL13, CCL20, IL24, IL22, etc.), while good responders are characterized by elevated expression of platelets (CD31 and PECAM), coagulation (CD74, F13A1), and complement classic pathway (C1QB, C1R, C1QBP, C1S, C2) markers, together with extracellular matrix proteins (COL3A1, FN1,. Interestingly the T-cell response is also different between groups. Poor responding lesions have increased Th1 and Th2 component, but good responders have an increased Th17 component. Interestingly, the checkpoint inhibitor of the immune response PD1 (PDCD1) is increased ~10 fold in poor responders. This preliminary study using a novel experimental approach revealed differences in the immune response in GCTs associated with clinical response to denosumab. The increased activity of checkpoint inhibitor PD1 in poor responders to denosumab treatment may have implications for therapy, raising the potential to investigate immunotherapy as is currently used in other neoplasms. Further validation using a larger independent cohort will be required but these results could potentially identify the patients who would most benefit from denosumab therapy

Aims

Factors associated with high mortality rates in geriatric hip fracture patients are frequently unmodifiable. Time to surgery, however, might be a modifiable factor of interest to optimize clinical outcomes after hip fracture surgery. This study aims to determine the influence of postponement of surgery due to non-medical reasons on clinical outcomes in acute hip fracture surgery.

Literature surrounding artificial intelligence (AI)-related applications for hip and knee arthroplasty has proliferated. However, meaningful advances that fundamentally transform the practice and delivery of joint arthroplasty are yet to be realized, despite the broad range of applications as we continue to search for meaningful and appropriate use of AI. AI literature in hip and knee arthroplasty between 2018 and 2021 regarding image-based analyses, value-based care, remote patient monitoring, and augmented reality was reviewed. Concerns surrounding meaningful use and appropriate methodological approaches of AI in joint arthroplasty research are summarized. Of the 233 AI-related orthopaedics articles published, 178 (76%) constituted original research, while the rest consisted of editorials or reviews. A total of 52% of original AI-related research concerns hip and knee arthroplasty (n = 92), and a narrative review is described. Three studies were externally validated. Pitfalls surrounding present-day research include conflating vernacular (“AI/machine learning”), repackaging limited registry data, prematurely releasing internally validated prediction models, appraising model architecture instead of inputted data, withholding code, and evaluating studies using antiquated regression-based guidelines. While AI has been applied to a variety of hip and knee arthroplasty applications with limited clinical impact, the future remains promising if the question is meaningful, the methodology is rigorous and transparent, the data are rich, and the model is externally validated. Simple checkpoints for meaningful AI adoption include ensuring applications focus on: administrative support over clinical evaluation and management; necessity of the advanced model; and the novelty of the question being answered.

Cite this article: Bone Joint J 2022;104-B(12):1292–1303.

Aims. We aimed to develop a gene signature that predicts the occurrence of postmenopausal osteoporosis (PMOP) by studying its genetic mechanism. Methods. Five datasets were obtained from the Gene Expression Omnibus database. Unsupervised consensus cluster analysis was used to determine new PMOP subtypes. To determine the central genes and the core modules related to PMOP, the weighted gene co-expression network analysis (WCGNA) was applied. Gene Ontology enrichment analysis was used to explore the biological processes underlying key genes. Logistic regression univariate analysis was used to screen for statistically significant variables. Two algorithms were used to select important PMOP-related genes. A logistic regression model was used to construct the PMOP-related gene profile. The receiver operating characteristic area under the curve, Harrell’s concordance index, a calibration chart, and decision curve analysis were used to characterize PMOP-related genes. Then, quantitative real-time polymerase chain reaction (qRT-PCR) was used to verify the expression of the PMOP-related genes in the gene signature. Results. We identified three PMOP-related subtypes and four core modules. The muscle system process, muscle contraction, and actin filament-based movement were more active in the hub genes. We obtained five feature genes related to PMOP. Our analysis verified that the gene signature had good predictive power and applicability. The outcomes of the GSE56815 cohort were found to be consistent with the results of the earlier studies. qRT-PCR results showed that RAB2A and FYCO1 were amplified in clinical samples. Conclusion. The PMOP-related gene signature we developed and verified can accurately predict the risk of PMOP in patients. These results can elucidate the molecular mechanism of RAB2A and FYCO1 underlying PMOP, and yield new and improved treatment strategies, ultimately helping PMOP monitoring. Cite this article: Bone Joint Res 2022;11(8):548–560

Aims

The aim of this study was to review the current evidence surrounding curve type and morphology on curve progression risk in adolescent idiopathic scoliosis (AIS).

Aims

Total hip arthroplasties (THAs) are performed by surgeons at various stages in training with varying levels of supervision, but we do not know if this is safe practice with comparable outcomes to consultant-performed THA. Our aim was to examine the association between surgeon grade, the senior supervision of trainees, and the risk of revision following THA.