Assessing retinal vein occlusion based on color fundus photographs using neural understanding network (NUN).

OBJECTIVE: To develop and validate a novel deep learning architecture to classify retinal vein occlusion (RVO) on color fundus photographs (CFPs) and reveal the image features contributing to the classification. METHODS: The neural understanding network (NUN) is formed by two components: (1) convolutional neural network (CNN)-based feature extraction and (2) graph neural networks (GNN)-based feature understanding. The CNN-based image features were transformed into a graph representation to encode and visualize long-range feature interactions to identify the image regions that significantly contributed to the classification decision. A total of 7062 CFPs were classified into three categories: (1) no vein occlusion ("normal"), (2) central RVO, and (3) branch RVO. The area under the receiver operative characteristic (ROC) curve (AUC) was used as the metric to assess the performance of the trained classification models. RESULTS: The AUC, accuracy, sensitivity, and specificity for NUN to classify CFPs as normal, central occlusion, or branch occlusion were 0.975 (± 0.003), 0.911 (± 0.007), 0.983 (± 0.010), and 0.803 (± 0.005), respectively, which outperformed available classical CNN models. CONCLUSION: The NUN architecture can provide a better classification performance and a straightforward visualization of the results compared to CNNs.

Do preoperative intra-articular corticosteroid and hyaluronic acid injections affect time to total joint arthroplasty?

INTRODUCTION: Intra-articular corticosteroid (CSI) or hyaluronic acid (HAI) injections alleviate symptoms of osteoarthritis in patients who may be candidates for total hip or total knee arthroplasty (THA/TKA). However, their effect on time to total joint arthroplasty (TJA) and complications remains uncertain. We sought to evaluate (1) delay in time to surgery for patients receiving injections prior to THA/TKA (2) incidence of patients that receive injections, (3) type and number of injections, and (4) compare complication rates between patients with and without injections. METHODS: We retrospectively reviewed 3340 consecutive TJA (1770 THA and 1570 TKA). Patients were divided into two cohorts depending if they received preoperative intra-articular injection or not. We identified dates of first clinic presentation and index surgery, injection type, total administered, and 90-day complications, including periprosthetic joint infection. RESULTS: 150/1770 THA and 192/1570 TKA patients received injections (8.5%vs.12.2%,p = 0.0004). Time from first presentation to clinic to TJA was significantly greater in patients receiving injections [12.4 ± 11 months vs.7.3 ± 10.7,p < 0.001 for THA; 20.0 ± 17.4 months vs.11.6 ± 15.4,p < 0.001 for TKA]. This delay in time was greater in TKA versus THA (8.4 months vs.5.1,p < 0.001). TKA patients had a higher incidence of receiving HAI versus THA patients (9%vs.0.6%,p < 0.0001). There were no differences in overall complication profiles (p = 0.19 for THA, p = 0.3 for TKA). CONCLUSION: Injections are associated with an increased time to TJA by a statistically significant amount, however its clinical significance is debatable. Injections are safe if administered at least three months preoperatively. If patients present with appropriate surgical indications and are ready, we do not recommend intra-articular injections to delay surgery.