Background: A critical part in preoperative planning for revision arthroplasty surgery involves the identification of the failed implant. Using a predictive artificial neural network (ANN) model, the objectives of this study were: (1) to develop a machine-learning algorithm using operative big data to identify an implant from a radiograph; and (2) to compare algorithms that optimise accuracy in a timely fashion. Methods: Using 2116 postoperative anteroposterior (AP) hip radiographs of total hip arthroplasties from 2002 to 2019, 10 artificial neural networks were modeled and trained to classify the radiograph according to the femoral stem implanted. Stem brand and model was confirmed with 1594 operative reports. Model performance was determined by classification accuracy toward a random 706 AP hip radiographs, and again on a consecutive series of 324 radiographs prospectively collected over 2019. Results: The Dense-Net 201 architecture outperformed all others with 100.00% accuracy in training data, 95.15% accuracy on validation data, and 91.16% accuracy in the unique prospective series of patients. This outperformed all other models on the validation ( p < 0.0001) and novel series ( p < 0.0001). The convolutional neural network also displayed the probability (confidence) of the femoral stem classification for any input radiograph. This neural network averaged a runtime of 0.96 (SD 0.02) seconds for an iPhone 6 to calculate from a given radiograph when converted to an application. Conclusions: Neural networks offer a useful adjunct to the surgeon in preoperative identification of the prior implant.

Objective: Abdominal aortic aneurysm (AAA) is a chronic inflammatory disease. Studies of human aneurysm tissue demonstrate dense inflammatory cell infiltrates with CD4 þ T cells predominating. Regulatory T cells (Tregs) play an important role in inhibiting pro-inflammatory T cell proliferation, therefore, limiting collateral tissue destruction. The aim of this study was to investigate whether ex vivo augmentation of human Tregs attenuates aneurysm formation in humanized murine model of AAA. Methods: Circulating Treg population in AAA patients and age-and gender-matched controls were determined by real-time polymerase chain reaction and flow cytometry. To create humanized murine model of AAA, irradiated Rag1deficient (Rag1 -/-) mice, without mature T lymphocytes, at 7 weeks of age were given 5 Â 10 6 of human CD4 þ T cells intraperitoneally. Then the mice underwent CaCl 2 aneurysm induction. Aortic diameters were measured before and at 6 weeks after aneurysm induction. Aortic tissue was collected for histology and protein extraction. Verhoeff-Van Gieson stain was used for staining elastic fiber. CD4 þ T cells in the aortic tissue were detected by immunohistochemical staining.Results: In human peripheral blood mononuclear cells, the proportion of Tregs are decreased in AAA patients compared with matched control patients with significant vascular disease. We first validated the role of Tregs in the CaCl 2 model of AAA. To determine the role of human T cells in AAA formation, Rag1 -/mice, resistant to CaCl 2 -aneurysm induction, were transplanted with human CD4 þ T cells. Human CD4 þ T cells were able to drive aneurysm formation in Rag1 -/mice. We show that ex vivo augmentation of human Tregs by interleukin-2 resulted in decreased aneurysm progression.Conclusions: These data suggest that the ex vivo expansion of human Tregs may be a potential therapeutic strategy for inhibiting progression of AAA.

Aims Several radiological methods of measuring anteversion of the acetabular component after total hip arthroplasty (THA) have been described. These are limited by low reproducibility, are less accurate than CT 3D reconstruction, and are cumbersome to use. These methods also partly rely on the identification of obscured radiological borders of the component. We propose two novel methods, the Area and Orthogonal methods, which have been designed to maximize use of readily identifiable points while maintaining the same trigonometric principles. Patients and Methods A retrospective study of plain radiographs was conducted on 160 hips of 141 patients who had undergone primary THA. We compared the reliability and accuracy of the Area and Orthogonal methods with two of the current leading methods: those of Widmer and Lewinnek, respectively. Results The 160 anteroposterior pelvis films revealed that the proposed Area method was statistically different from those described by Widmer and Lewinnek (p < 0.001 and p = 0.004, respectively). They gave the highest inter- and intraobserver reliability (0.992 and 0.998, respectively), and took less time (27.50 seconds (sd 3.19); p < 0.001) to complete. In addition, 21 available CT 3D reconstructions revealed the Area method achieved the highest Pearson’s correlation coefficient (r = 0.956; p < 0.001) and least statistical difference (p = 0.704) from CT with a mean within 1° of CT-3D reconstruction between ranges of 1° to 30° of measured radiological anteversion. Conclusion Our results support the proposed Area method to be the most reliable, accurate, and speedy. They did not support any statistical superiority of the proposed Orthogonal method to that of the Widmer or Lewinnek method. Cite this article: Bone Joint J 2019;101-B:1042–1049.

Objectives: Far cortical locking (FCL) constructs have been shown to increase axial interfragmentary displacement while limiting shear and have been specifically recommended in the treatment of distal femur fractures. However, there is no available data regarding their mechanical behavior within the range of bridge spans typically used for comminuted distal femur fractures. This biomechanical study of distal femur locked plate fixation assessed 4 methods of diaphyseal fixation for associated axial and shear displacement at bridge spans typically used in clinical practice. Methods: Distal femur locking plates were used to bridge simulated fractures in femur surrogates with 4 different methods of diaphyseal fixation (bicortical locking, bicortical nonlocking, near cortical locking, and FCL). Axial and shear displacement were assessed at 5 different bridge spans for each fixation method. Results: Diaphyseal fixation type was associated with the amount of shear (P = 0.04), but not the amount of axial displacement (P = 0.39). Specifically, FCL constructs demonstrated greater shear than bicortical locking (median 4.57 vs. 2.94 mm, P = 0.02) and bicortical nonlocking (median 4.57 vs. 3.41 mm, P = 0.02) constructs. Conclusions: Unexpectedly, FCL constructs demonstrated greater shear than bicortical locking and nonlocking constructs and similar axial displacement for all fixation methods. Bridge span had a dominant effect on displacement that interacted negatively with more flexible FCL diaphyseal fixation. Potentially interactive construct features are best studied in concert. Given the complexity of these relationships, computational modeling will likely play an integral role in future mechanotransduction research.