Background and purpose — Recent advances in artificial intelligence (deep learning) have shown remarkable performance in classifying non-medical images, and the technology is believed to be the next technological revolution. So far it has never been applied in an orthopedic setting, and in this study we sought to determine the feasibility of using deep learning for skeletal radiographs.Methods — We extracted 256,000 wrist, hand, and ankle radiographs from Danderyd’s Hospital and identified 4 classes: fracture, laterality, body part, and exam view. We then selected 5 openly available deep learning networks that were adapted for these images. The most accurate network was benchmarked against a gold standard for fractures. We furthermore compared the network’s performance with 2 senior orthopedic surgeons who reviewed images at the same resolution as the network.Results — All networks exhibited an accuracy of at least 90% when identifying laterality, body part, and exam view. The final accuracy for fractures was estimated at 83% for the best performing network. The network performed similarly to senior orthopedic surgeons when presented with images at the same resolution as the network. The 2 reviewer Cohen’s kappa under these conditions was 0.76.Interpretation — This study supports the use for orthopedic radiographs of artificial intelligence, which can perform at a human level. While current implementation lacks important features that surgeons require, e.g. risk of dislocation, classifications, measurements, and combining multiple exam views, these problems have technical solutions that are waiting to be implemented for orthopedics.

Background and purpose The effects of patient-related and technical factors on the risk of revision due to dislocation after primary total hip arthroplasty (THA) are only partly understood. We hypothesized that increasing the femoral head size can reduce this risk, that the lateral surgical approach is associated with a lower risk than the posterior and minimally invasive approaches, and that gender and diagnosis influence the risk of revision due to dislocation.Patients and methods Data on 78,098 THAs in 61,743 patients performed between 2005 and 2010 were extracted from the Swedish Hip Arthroplasty Register. Inclusion criteria were a head size of 22, 28, 32, or 36 mm, or the use of a dual-mobility cup. The covariates age, sex, primary diagnosis, type of surgical approach, and head size were entered into Cox proportional hazards models in order to calculate the adjusted relative risk (RR) of revision due to dislocation, with 95% confidence intervals (CI).Results After a mean follow-up of 2.7 (0–6) years, 399 hips (0.5%) had been revised due to dislocation. The use of 22-mm femoral heads resulted in a higher risk of revision than the use of 28-mm heads (RR = 2.0, CI: 1.2–3.3). Only 1 of 287 dual-mobility cups had been revised due to dislocation. Compared with the direct lateral approach, minimally invasive approaches were associated with a higher risk of revision due to dislocation (RR = 4.2, CI: 2.3–7.7), as were posterior approaches (RR = 1.3, CI: 1.1–1.7). An increased risk of revision due to dislocation was found for the diagnoses femoral neck fracture (RR = 3.9, CI: 3.1–5.0) and osteonecrosis of the femoral head (RR = 3.7, CI: 2.5–5.5), whereas women were at lower risk than men (RR = 0.8, CI: 0.7–1.0). Restriction of the analysis to the first 6 months after the index procedure gave similar risk estimates.Interpretation Patients with femoral neck fracture or osteonecrosis of the femoral head are at higher risk of dislocation. Use of the minimally invasive and posterior approaches also increases this risk, and we raise the question of whether patients belonging to risk groups should be operated using lateral approaches. The use of femoral head diameters above 28 mm or of dual-mobility cups reduced this risk in a clinically relevant manner, but this observation was not statistically significant.

After the hand, the foot was the most frequently symptomatic joint complex at the start of the disease, but also during active medical treatment. The foot caused walking disability in three-quarters of the cases and-4 times as often as the knee or the hip-it was the only joint to subjectively impair gait.