Complex system anomaly detection via learnable temporal-spatial graph with degradation tendency segmentation.

To guarantee the safety and reliability of equipment operation, such as liquid rocket engine (LRE), carrying out system-level anomaly detection (AD) is crucial. However, current methods ignore the prior knowledge of mechanical system itself, and seldom unite the observations with the inherent relation in data tightly. Meanwhile, they neglect the weakness and nonindependence of system-level anomaly which is different from component fault. To overcome above limitations, we propose a separate reconstruction framework using worsened tendency for system-level AD. To prevent anomalous feature being attenuated, we first propose to divide single sample into two equal-length parts along the temporal dimension. And we maximize the mean maximum discrepancy (MMD) between feature segments to force encoders to learn normal features with different distributions. Then, to fully explore the multivariate time series, we model temporal-spatial dependence by temporal convolution and graph attention. Besides, a joint graph learning strategy is proposed to handle prior knowledge and data characteristics simultaneously. Finally, the proposed method is evaluated on two real multi-sensor datasets from LRE and the results demonstrate the effectiveness and potential of the proposed method on system-level AD.

Comparison of multiple biomarkers for mortality prediction in patients with acute heart failure of ischemic and nonischemic etiology.

AIM: To investigate the prognosis of soluble ST2 (sST2), galectin-3 and N-terminal pro-B-type natriuretic peptide (NT-proBNP) for death related to ischemic and nonischemic etiology of acute heart failure (HF). METHODS: The associations between biomarkers and death were determined in 1020 patients admitted to hospital with acute HF. RESULTS: During 1-year follow-up, 162 patients died. Multivariable regression analysis showed that the hazard ratios of sST2 and NT-proBNP for 1-year all-cause death was similar and remained significant between ischemic and nonischemic HF patients. However, galectin-3 was not significantly associated with death when sST2 and NT-proBNP were incorporated into model in ischemic HF patients. CONCLUSION: There is no etiology dependent prognostic ability of NT-proBNP or ST2 in patients with acute HF, but for galectin-3 there is no added prognostic ability in ischemic HF.

Hemodialysis is associated with higher serum FGF23 level when compared with peritoneal dialysis.

BACKGROUND: Considering the growing relevance of fibroblast growth factor-23 (FGF23) and increased cardiovascular mortality in dialysis population, an analysis was performed to assess the influence of dialysis modality (peritoneal dialysis and hemodialysis) on level of FGF23. METHODS: A cross-sectional study was performed in 80 continuous ambulatory peritoneal dialysis (CAPD) and 65 hemodialysis (HD) patients without residual renal function. Levels of calcium, phosphate, parathyroid hormone and FGF23 were measured, and their correlations were analyzed. Data on demographics, dialysis modality and FGF23 level were also analyzed. RESULTS: A significant correlation was found between FGF23 and serum calcium, serum phosphate and dialysis vintage in dialysis patients. Level of FGF23 was significantly higher in hemodialysis patients than that in peritoneal dialysis population. Multivariable regression revealed that, compared to CAPD, hemodialysis was found to be a predictor for higher FGF23 level, which was independent of serum calcium and phosphate level (P < 0.05). CONCLUSIONS: These findings demonstrate that FGF23 levels are significantly higher in hemodialysis patients than that in peritoneal dialysis patients. We demonstrate an important association between dialysis modality (HD vs CAPD) and higher FGF23, independent of classical determinants (serum calcium and phosphate level).