Search

In practical engineering scenarios with limited labeled samples, conventional semi-supervised diagnostic methods face challenges in achieving satisfactory identification outcomes. To address the aforementioned issues, this paper introduces an incremental semi-supervised learning (ISL) approach based on boosting efficient attention (BEA) assisted cyclic adversarial auto-encoder (CAAE), referred to as BEA-CAAE. The CAAE enhances unsupervised feature representation by simultaneously constraining the distribution of encoded features and aligning elements of the reconstructed samples through a cyclic encoding strategy. The BEA improves classical attention weights' activation strength to better capture vital information, thereby boosting the feature extraction capabilities of both CAAE and the classifier. The ISL employs a stepwise pseudo-label propagation strategy to incrementally filter high-confidence samples, enhancing sample and label utilization, and improving diagnostic accuracy under low label rate conditions. Experiments conducted on multiple test rigs with simple structures as well as large-scale rotating components test rigs that mimic real-world engineering conditions have demonstrated that the proposed method exhibits a significant advantage over existing semi-supervised fault diagnosis approaches in terms of fault diagnosis accuracy and generalization capability, especially under low label rate conditions. [ABSTRACT FROM AUTHOR]

• A semi-supervised ensemble framework termed ADAE-LFDM is proposed for fault diagnosis in extremely limited labels scenarios. • An adversarial decoupled auto-encoder network for fault feature decoupling is designed to enhance the representation capabilities of fault feature. • The low-dimensional feature distance metric is designed to enable accurate recognition of new data, which can mitigate the disadvantage imposed by extremely limited labels. • Experiments in extremely limited labels scenarios are designed on bearing and gear datasets to verify the superiority of the proposed methods. Intelligent fault diagnosis can enhance the reliability of mechanical equipment. However, only a few labels are available in a large amount of fault data due to high labeling costs in practical engineering. The fault recognition capability of existing semi-supervised diagnosis methods is severely insufficient with limited labels, especially with extremely limited labels that only a single labeled sample available per fault type. To address this issue, a novel semi-supervised ensemble fault diagnosis framework termed ADAE-LFDM is proposed based on adversarial decoupled auto-encoder (ADAE) and low-dimensional feature distance metric (LFDM). Firstly, the locally selective combination in parallel outlier ensembles (LSCP) method is introduced to efficiently separate normal and fault samples. Subsequently, an ADAE with branching structure and latent space feature regularization strategy is proposed to decouple and capture the fault feature. Finally, a LFDM strategy that contains feature dimensionality reduction, and centroid-based metric is performed to achieve high-accuracy fault diagnosis. Experimental results based on two rotating machinery datasets have demonstrated that the proposed method achieves a diagnostic accuracy of over 97 % when there is only a single labeled sample available per fault type, and an average diagnostic accuracy of 85 % under cross-operating condition, showing the superiority compared to other methods. [ABSTRACT FROM AUTHOR]

Skeletal muscle atrophy in chronic kidney disease (CKD) leads to a decline in quality of life and increased risk of morbidity and mortality. We have obtained evidence that oxidative stress is essential in the progression of CKD-related muscle atrophy. Whether Saikosaponin A and D, two emerging antioxidants extracted from Bupleurum chinense DC, alleviate muscle atrophy remains to be further studied. The purpose of this study was to investigate the effects and mechanisms of these two components on CKD complicated with muscle atrophy. In this research, muscle dystrophy model was established using 5/6 nephrectomized mice in vivo and in vitro with Dexamethasone (Dex)-managed C2C12 myotubes. The results of RNA-sequencing showed that exposure to Dex affected the antioxidant activity, catalytic activity and enzyme regulator activity of C2C12 cells. According to KEGG analysis, the largest numbers of differentially expressed genes detected were enriched in the PI3K/AKT pathway. In vivo , Saikosaponin A and D remain renal function, cross-section size, fiber-type composition and anti-inflammatory ability. These two components suppressed the expression of MuRF-1 and enhanced the expression of MyoD and Dystrophin. In addition, Saikosaponin A and D maintained redox balance by increasing the activities of antioxidant enzymes while inhibiting the excessive accumulation of reactive oxygen species. Furthermore, Saikosaponin A and D stimulated PI3K/AKT and its downstream Nrf2 pathway in CKD mice. The effects of Saikosaponin A and D on increasing the inner diameter of C2C12 myotube, reducing oxidative stress and enhancing expression of p-AKT, p-mTOR, p70S6K, Nrf2 and HO-1 proteins were observed in vitro. Importantly, we verified that these protective effects could be significantly reversed by inhibiting PI3K and knocking out Nrf2. In summary, Saikosaponin A and D improve CKD-induced muscle atrophy by reducing oxidative stress through the PI3K/AKT/Nrf2 pathway. [Display omitted] [ABSTRACT FROM AUTHOR]

Voltage-dependent anion-selective channel protein 1 (VDAC1) is the most abundant protein in the mitochondrial outer membrane and plays a crucial role in the control of hepatocellular carcinoma (HCC) progress. Our previous research found that cytosolic molecular chaperone heat shock protein 90 (Hsp90) interacted with VDAC1, but the effect of the C-terminal and N-terminal domains of Hsp90 on the formation of VDAC1 oligomers is unclear. In this study, we focused on the effect of the C-terminal domain of Hsp90 on VDAC1 oligomerization, ubiquitination, and VDAC1 channel activity. We found that Hsp90 C-terminal domain inhibitor Novobiocin promoted VDAC1 oligomerization, release of cytochrome c, and activated mitochondrial apoptosis pathway. Atomic coarse particle modeling simulation revealed C-terminal domain of Hsp90α stabilized VDAC1 monomers. The purified VDAC1 was reconstituted into a planar lipid bilayer, and electrophysiology experiments of patch clamp showed that the Hsp90 C-terminal inhibitor Novobiocin increased VDAC1 channel conductance via promoting VDAC1 oligomerization. The mitochondrial ubiquitination proteomics results showed that VDAC1 K274 mono-ubiquitination was significantly decreased upon Novobiocin treatment. Site-directed mutation of VDAC1 (K274R) weakened Hsp90α-VDAC1 interaction and increased VDAC1 oligomerization. Taken together, our results reveal that Hsp90 C-terminal domain inhibition promotes VDAC1 oligomerization and VDAC1 channel conductance by decreasing VDAC1 K274 mono- ubiquitination, which provides a new perspective for mitochondria-targeted therapy of HCC. [Display omitted] [ABSTRACT FROM AUTHOR]