Automated identification of arthroplasty implants could aid in pre-operative planning and is a task which could be facilitated through artificial intelligence (AI) and deep learning. The purpose of this study was to develop and test the performance of a deep learning system (DLS) for automated identification and classification of knee arthroplasty (KA) on radiographs. We collected 237 AP knee radiographs with equal proportions of native knees, total KA (TKA), and unicompartmental KA (UKA), as well as 274 radiographs with equal proportions of Smith & Nephew Journey and Zimmer NexGen TKAs. Data augmentation was used to increase the number of images available for DLS development. These images were used to train, validate, and test deep convolutional neural networks (DCNN) to 1) detect the presence of TKA; 2) differentiate between TKA and UKA; and 3) differentiate between the 2 TKA models. Receiver operating characteristic (ROC) curves were generated with area under the curve (AUC) calculated to assess test performance.Introduction
Methods
Highly cross-linked polyethylene (XLPE) inserts have shown significant improvements in decreasing wear and osteolysis in total hip arthroplasty [1]. In contrast to that, XLPE has not shown to reduce wear or aseptic loosening in total knee arthroplasty [2,3,4]. One major limitation is that current wear testing in vitro is mainly focused on abrasive-adhesive wear due to level walking test conditions and does not reflect “delamination” as an essential clinical failure mode [5,6]. The objective of our study was to use a highly demanding daily activities wear simulation to evaluate the delamination risk of polyethylene materials with and without vitamin E stabilisation. A cruciate retaining fixed bearing TKA design (Columbus® CR) with artificially aged polyethylene knee bearings (irradiation 30 & 50 kGy) blended with and without 0.1% vitamin E was used under medio-lateral load distribution and soft tissue restrain simulation. Daily patient activities measured by Bergmann et al. [7] in vivo, were applied for 5 million knee wear cycles in a combination of 40% stairs up, 40 % stairs down, 10% level walking, 8% chair raising and 2% deep squatting with up to 100° flexion [8] (Fig. 1). The specimens were evaluated for gravimetric wear and analysed for abrasive-adhesive and delamination wear modes.INTRODUCTION
MATERIALS & METHODS
