Lightweight Attentive Graph Neural Network with Conditional Random Field for Diagnosis of Anterior Cruciate Ligament Tear.

Anterior cruciate ligament (ACL) tears are prevalent orthopedic sports injuries and are difficult to precisely classify. Previous works have demonstrated the ability of deep learning (DL) to provide support for clinicians in ACL tear classification scenarios, but it requires a large quantity of labeled samples and incurs a high computational expense. This study aims to overcome the challenges brought by small and imbalanced data and achieve fast and accurate ACL tear classification based on magnetic resonance imaging (MRI) of the knee. We propose a lightweight attentive graph neural network (GNN) with a conditional random field (CRF), named the ACGNN, to classify ACL ruptures in knee MR images. A metric-based meta-learning strategy is introduced to conduct independent testing through multiple node classification tasks. We design a lightweight feature embedding network using a feature-based knowledge distillation method to extract features from the given images. Then, GNN layers are used to find the dependencies between samples and complete the classification process. The CRF is incorporated into each GNN layer to refine the affinities. To mitigate oversmoothing and overfitting issues, we apply self-boosting attention, node attention, and memory attention for graph initialization, node updating, and correlation across graph layers, respectively. Experiments demonstrated that our model provided excellent performance on both oblique coronal data and sagittal data with accuracies of 92.94% and 91.92%, respectively. Notably, our proposed method exhibited comparable performance to that of orthopedic surgeons during an internal clinical validation. This work shows the potential of our method to advance ACL diagnosis and facilitates the development of computer-aided diagnosis methods for use in clinical practice.

Revealing the Effect of Surface Composition on Multiwalled Carbon Nanotubes Supported Pt-Fe Alloy Electrocatalysts for Methanol Oxidation Performance.

The designs of efficient and inexpensive Pt-based catalysts for methanol oxidation reaction (MOR) are essential to boost the commercialization of direct methanol fuel cells. Here, the highly catalytic performance PtFe alloys supported on multiwalled carbon nanotubes (MWCNTs) decorating nitrogen-doped carbon (NC) have been successfully prepared via co-engineering of the surface composition and electronic structure. The Pt1 Fe3 @NC/MWCNTs catalyst with moderate Fe3+ feeding content (0.86 mA/mgPt ) exhibits 2.26-fold enhancement in MOR mass activity compared to pristine Pt/C catalyst (0.38 mA/mgPt ). Furthermore, the CO oxidation initial potential of Pt1 Fe3 @NC/MWCNTs catalyst is lower relative to Pt/C catalyst (0.71 V and 0.80 V). Benefited from the optimal surface compositions, the anti-corrosion ability of MWCNT, strong electron interaction between PtFe alloys and MWCNTs and the N-doped carbon (NC) layer, the Pt1 Fe3 @NC/MWCNTs catalyst presents an improved MOR performance and anti-CO poisoning ability. This study would open up new perspective for designing efficient electrocatalysts for the DMFCs field.

TNF-α augments CXCL10/CXCR3 axis activity to induce Epithelial-Mesenchymal Transition in colon cancer cell.

Chronic inflammation-induced metastases have long been regarded as one of the significant obstacles in treating cancer. Tumor necrosis factor-α (TNF-α), a main inflammation mediator within tumor microenvironment, affects tumor development by inducing multiple chemokines to establish a complex network. Recent reports have revealed that CXCL10/CXCR3 axis affects cancer cells invasiveness and metastases, and Epithelial-mesenchymal transition (EMT) is the main reason for frequent proliferation and distant organ metastases of colon cancer (CC) cells, However, it is unclear whether TNF-α- mediated chronic inflammation can synergically enhance EMT-mediated CC metastasis through promoting chemokine expression. According to this study, TNF-α activated the PI3K/Akt and p38 MAPK parallel signal transduction pathways, then stimulate downstream NF-κB pathway p65 into the nucleus to activate CXCL10 transcription. CXCL10 enhanced the metastases of CC-cells by triggering small GTPases such as RhoA and cdc42. Furthermore, overexpression of CXCL10 significantly enhanced tumorigenicity and mobility of CC cells in vivo. We further clarified that CXCL10 activated the PI3K/Akt pathway through CXCR3, resulting in suppression of GSK-3ß phosphorylation and leading to upregulation of Snail expression, thereby regulating EMT in CC cells. These outcomes lay the foundation for finding new targets to inhibit CC metastases.

High-dose dexamethasone vs prednisone for treatment of adult immune thrombocytopenia: a prospective multicenter randomized trial.

This study compared the efficacy and safety of high-dose dexamethasone (HD-DXM) and conventional prednisone (PDN) on the largest cohort to date as first-line strategies for newly diagnosed adult primary immune thrombocytopenia (ITP). Patients enrolled were randomized to receive DXM 40 mg/d for 4 days (n = 95, nonresponders received an additional 4-day course of DXM) or prednisone 1.0 mg/kg daily for 4 weeks and then tapered (n = 97). One or 2 courses of HD-DXM resulted in a higher incidence of overall initial response (82.1% vs 67.4%, P = .044) and complete response (50.5% vs 26.8%, P = .001) compared with prednisone. Time to response was shorter in the HD-DXM arm (P < .001), and a baseline bleeding score ≥8 was associated with a decreased likelihood of initial response. Sustained response was achieved by 40.0% of patients in the HD-DXM arm and 41.2% in the PDN arm (P = .884). Initial complete response was a positive indicator of sustained response, whereas presence of antiplatelet autoantibodies was a negative indicator. HD-DXM was generally tolerated better. We concluded that HD-DXM could be a preferred corticosteroid strategy for first-line management of adult primary ITP. This study is registered at www.clinicaltrials.gov as #NCT01356511.

The interaction between bamboo mosaic virus replication protein and coat protein is critical for virus movement in plant hosts.

Bamboo mosaic virus (BaMV) is a positive-sense RNA virus belonging to the genus Potexvirus. Open reading frame 1 (ORF1) encodes the viral replication protein that consists of a capping enzyme domain, a helicase-like domain (HLD), and an RNA-dependent RNA polymerase domain from the N to C terminus. ORF5 encodes the viral coat protein (CP) required for genome encapsidation and the virus movement in plants. In this study, application of a yeast-two hybrid assay detected an interaction between the viral HLD and CP. However, the interaction did not affect the NTPase activity of the HLD. To identify the critical amino acids of CP interacting with the HLD, a random mutational library of CP was created using error-prone PCR, and the mutations adversely affecting the interaction were screened by a bacterial two-hybrid system. As a result, the mutations A209G and N210S in CP were found to weaken the interaction. To determine the significance of the interaction, the mutations were introduced into a BaMV infectious clone, and the mutational effects on viral replication, movement, and genome encapsidation were investigated. There was no effect on accumulations of BaMV CP and genomic RNAs within protoplasts; however, the virus cell-to-cell movement in plants was restricted. Sequence alignment revealed that A209 of BaMV CP is conserved in many potexviruses. Mutation of the corresponding residue in Foxtail mosaic virus CP also reduced the viral HLD-CP interaction and restricted the virus movement, suggesting that interaction between CP and a widely conserved HLD in the potexviral replication protein is crucial for viral trafficking through plasmodesmata.