Search

Creating an effective deep learning technique for accurately diagnosing leak signals across diverse environments is crucial for integrating artificial intelligence (AI) into the power plant industry. We propose an automatic weight redistribution ensemble model based on transfer learning (TL) for detecting leaks in diverse power plant environments, overcoming the challenges of site-specific AI methods. This innovative model processes time series acoustic data collected from multiple homogeneous sensors located at different positions into three-dimensional root-mean-square (RMS) and frequency volume features, enabling accurate leak detection. Utilizing a TL-driven, two-stage learning process, we first train residual-network-based models for each domain using these preprocessed features. Subsequently, these models are retrained in an ensemble for comprehensive leak detection across domains, with control weight ratios finely adjusted through a softmax score-based approach. The experiment results demonstrate that the proposed method effectively distinguishes low-level leaks and noise compared to existing techniques, even when the data available for model training are very limited. [ABSTRACT FROM AUTHOR]

Background: Periprosthetic joint infection (PJI) is one of the most serious and debilitating complications that can occur after total joint arthroplasty. Therefore, early diagnosis and appropriate treatment are important for a good prognosis. Recently, molecular diagnostic methods have been widely used to detect the causative microorganisms of PJI sensitively and rapidly. The Multiplex Loop-Mediated Isothermal Amplification (LAMP) method eliminates the complex temperature cycling and delays caused by temperature transitions seen in polymerase chain reaction (PCR) methods, making it faster and easier to perform compared to PCR-based assays. Therefore, this study developed a multiplex LAMP assay for diagnosing bacterial PJI using LAMP technology and evaluated its analytical and clinical performance. Methods: We developed a multiplex LAMP assay for the detection of five bacteria: Staphylococcus aureus, Staphylococcus epidermidis, Streptococcus agalactiae, Pseudomonas aeruginosa, and Escherichia coli, frequently observed to be the causative agents of PJI. The method of analytical sensitivity and cross-reactivity were determined by spiking standard strains into the joint synovial fluid. The analytical sensitivity of the multiplex LAMP assay was compared with that of a quantitative real-time PCR (qPCR) assay. Clinical performance was evaluated using 20 joint synovial fluid samples collected from patients suspected of having bacterial PJI. Results: The analytical sensitivity of the gram-positive bacterial multiplex LAMP assay and qPCR were 105/104 CFU/mL, 103/103 CFU/mL, and 105/104 CFU/mL against S. agalactiae, S. epidermidis, and S. aureus, respectively. For P. aeruginosa and E. coli, the analytical sensitivity of the multiplex LAMP and qPCR assays were 105/104 and 106/104 CFU/mL, respectively. The multiplex LAMP assay detects target bacteria without cross-reacting with other bacteria, and exhibited 100% sensitivity and specificity in clinical performance evaluation. Conclusions: This multiplex LAMP assay can rapidly detect five high-prevalence bacterial species causing bacterial PJI, with excellent sensitivity and specificity, in less than 1 h, and it may be useful for the early diagnosis of PJI. [ABSTRACT FROM AUTHOR]

This study aimed to assess the safety and effectiveness of the highly cross-linked hyaluronic acid—LBSA0103—in patients with knee osteoarthritis (OA) as per the prescribing information (PI) in South Korea. A total of 3,140 subjects aged ≥19 years were enrolled in this post-marketing surveillance (PMS) study from 2013 to 2019. The subjects received one or two injections of LBSA0103. The median duration of follow-up was 308 days. Adverse events (AEs), adverse drug reactions (ADRs), and serious AEs (SAEs) were monitored. Effectiveness was evaluated based on an index of effectiveness in accordance with the guidelines established by the Ministry of Food and Drug Safety and using a 100-mm visual analog scale (VAS) for weight-bearing pain. Overall, 250 subjects (7.96%) experienced 292 AEs and of these, unexpected AEs occurred in 114 subjects (3.63% [95% CI: 3.00–4.35]). Injection site pain was the most frequent AE reported by 81 subjects (2.58% [95% confidence intervals (CI): 2.05–3.20]). One hundred subjects experienced 108 ADRs (3.18% [95% CI: 2.60, 3.86]) and 15 unexpected ADRs were experienced by 13 subjects (0.41% [95% CI: 0.22–0.71]). Seventeen subjects experienced 22 SAEs (0.54% [95% CI: 0.32–0.87]) during the entire PMS period, and all were considered "unlikely" related to the study drug. Most AEs were mild in terms of severity and resolved during the study period. LBSA0103 was also effective in relieving symptomatic pain in knee OA patients. The condition in more than 80% of the subjects was considered to be improved when assessed by the investigators. LBSA0103 resulted in a significant reduction in the mean VAS score at 12 weeks after the first and second injections (24.79 (± 20.55) mm and 17.63 (±12.31) mm, respectively; p<0.0001). In conclusion, LBSA0103, used for the treatment of knee OA in a real-world setting, was well tolerated, with an acceptable safety profile and consistent therapeutic effect. [ABSTRACT FROM AUTHOR]