Sensor Data Reconstruction for Dynamic Responses of Structures Using External Feedback of Recurrent Neural Network.

An event of sensor faults in sensor networks deployed in structures might result in the degradation of the structural health monitoring system and lead to difficulties in structural condition assessment. Reconstruction techniques of the data for missing sensor channels were widely adopted to restore a dataset from all sensor channels. In this study, a recurrent neural network (RNN) model combined with external feedback is proposed to enhance the accuracy and effectiveness of sensor data reconstruction for measuring the dynamic responses of structures. The model utilizes spatial correlation rather than spatiotemporal correlation by explicitly feeding the previously reconstructed time series of defective sensor channels back to the input dataset. Because of the nature of spatial correlation, the proposed method generates robust and precise results regardless of the hyperparameters set in the RNN model. To verify the performance of the proposed method, simple RNN, long short-term memory, and gated recurrent unit models were trained using the acceleration datasets obtained from laboratory-scaled three- and six-story shear building frames.

Automated Dimension Estimation of Steel Pipes Stacked at Construction Sites Using Euclidean Distances Calculated by Overlapping Segmentation.

Pipes are construction materials for water and sewage, air conditioning, firefighting, and gas facilities at construction sites. The quantification and identification of pipes stacked at construction sites are indispensable and, thus, are directly related to efficient process management. In this study, an automated CNN-based technique for estimating the diameter and thickness of the pipe in an image is proposed. The proposed method infers the thickness of the pipe through the difference by segmentation, by overlapping the inside and outside circles for a single pipe. When multiple pipes are included in the image, the inside and outside circles for the identical pipe are matched through the spatial Euclidean distance. The CNN models are trained using pipe images of various sizes to segment the pipe circles. An error of less than 7.8% for the outer diameter and 15% for the thickness is verified through execution with a series of 50 testing pipe images.

Multipoint Wave Measurement in Tuned Liquid Damper Using Laser Doppler Vibrometer and Stepwise Rotating Galvanometer Scanner.

Liquid dampers, such as tuned liquid dampers (TLDs), are employed to improve serviceability by reducing wind-affected building vibrations. In order to maximize the vibration suppression efficiency of the liquid damper, the tuning frequency of the liquid damper should match the natural frequency of the building. Experimental evaluation of the tuning frequency of a liquid damper performed in a factory prior to installation in a building is a critical task to ensure correct performance, and for this, multipoint measurement of the TLD is required. In this study, a novel liquid level measurement system combining Laser Doppler Vibrometer (LDV) and a stepwise rotating galvanometer scanner was developed to observe liquid sloshing in TLD. The proposed system can measure the liquid level at multiple points simultaneously with a single laser point. In the experimental phase, the liquid damper's natural frequency and mode shape are experimentally evaluated utilizing the developed system. The performance of the proposed system was verified by comparison with the video sensing system.

Predictive value of serum albumin-to-globulin ratio for incident chronic kidney disease: A 12-year community-based prospective study.

BACKGROUND: Inflammation plays a pivotal role in the pathogenesis of chronic kidney disease (CKD). Significant association between serum albumin-to-globulin (AG) ratio and inflammation led us to investigate the prognostic value of serum AG ratio for incident CKD. METHODS: The predictive value of serum AG ratio, white blood cell (WBC), and C-reactive protein (CRP) for CKD development was assessed in 8,057 non-CKD participants from a community-based, prospective cohort in Korea. Serum AG ratio was calculated by following equation: serum albumin (g/L)/[serum total protein (g/L)-serum albumin (g/L)]. Incident CKD was defined as estimated glomerular filtration rate <60 mL/min/1.73 m2 and/or proteinuria of more than 1+ on dipstick. RESULTS: Median serum AG ratio was 1.38 (interquartile range, 1.28-1.52). During a mean follow-up duration of 9.1±3.7 years, 1,732 participants (21.5%) developed CKD. In a multivariable Cox analysis, a low serum AG ratio was significantly associated with an increased risk of incident CKD (Q1, serum AG ratio <1.26: hazard ratio [HR] = 1.651, 95% confidence interval [CI] = 1.406-1.938, Q5 as reference; per 0.2 decrease, HR = 1.170, 95% CI = 1.109-1.234). Serum AG ratio was the only indicator to improve the predictability of CKD development (net reclassification index = 0.158, P <0.001; integrated discrimination improvement = 0.005, P <0.001), compared with WBC or CRP. CONCLUSIONS: This study demonstrates that low serum AG ratio is an independent predictor for CKD development and exhibits a stronger predictive value than other inflammatory markers. These findings suggest that determining serum AG ratio may be more valuable for predicting adverse kidney outcomes in non-CKD populations.

Effects of online hemodiafiltration on anemia and nutritional status in chronic hemodialysis patients.

BACKGROUND: Online hemodiafiltration (OL-HDF) offers considerable advantages in clearance of molecules of various sizes. However, evidence of clinical effects of OL-HDF is scarce in Korea. In this study, we investigated changes in laboratory values over more than 12 months after switching to OL-HDF. METHODS: Adult patients with end-stage renal disease undergoing hemodialysis (HD) were prospectively enrolled in a K-cohort (CRIS no. KCT0003281) from 6 tertiary hospitals in South Korea. We recruited 435 patients, 339 of whom were on HD at enrollment. One hundred eighty-two patients were followed for more than 24 months. Among them, 44 were switched to OL-HDF for more than 12 months without conversion to HD. We used a paired t test to compare baseline and 24-month follow-up results. RESULTS: The mean age of the subjects was 61.2 ± 12.2 years, and 62.6% were male. The baseline hemoglobin level was not significantly different between HD and OL-HDF group (10.61 ± 1.15 vs. 10.46 ± 1.03 g/dL, P = 0.437). However, the baseline serum protein and albumin levels were significantly lower in the OL-HDF group (6.82 ± 0.49 vs. 6.59 ± 0.48 g/dL, P = 0.006; 3.93 ± 0.28 vs. 3.73 ± 0.29 g/dL, P < 0.001). In patients switched to OL-HDF, levels of hemoglobin and serum albumin significantly increased (10.46 ± 1.03 vs. 11.08 ± 0.82 g/dL, P = 0.001; 3.73 ± 0.29 vs. 3.87 ± 0.30 g/dL, P = 0.001). The normalized protein catabolic rate decreased after 24 months, but the change was not significant (1.07 ± 0.25 vs. 1.03 ± 0.21 g/kg/day, P = 0.433). Although the dose of erythropoiesis-stimulating agent was lower in patients who converted to HDF, it was not significantly different (-115.7 ± 189.7 vs. -170.5 ± 257.1 P = 0.206). CONCLUSION: OL-HDF treatment over more than 12 months was associated with no harmful effects on anemia and nutritional status.