Xiaoping Huang

Background: Rapid pathogen identification is essential for guiding timely and appropriate antimicrobial therapy in severe pulmonary infections. Microbiological rapid on-site evaluation (M-ROSE) can provide preliminary etiological information at the bedside, but its interpretation is labor intensive and highly dependent on experienced clinicians or microbiologists. This study aimed to develop and evaluate a deep learning framework for bacterial identification in sputum M-ROSE smears under limited annotated data conditions. Methods: A total of 161 Gram-stained sputum M-ROSE images containing Acinetobacter baumannii, Klebsiella pneumoniae, and Pseudomonas aeruginosa were collected using a digital scanner. Thirty annotated images were used to construct Dataset N, which was divided into a training set and a validation set at a ratio of 8:2. A deep learning model based on the YOLOv6 architecture was developed using transfer learning and few-shot learning. In addition to conventional data augmentation, two customized strategies—rotation cutting and image fusion—were introduced to enhance the detection of extremely small bacterial targets. The remaining 131 images were used as an independent testing set. Model performance was evaluated using recall, precision, F1 score, mean average precision (mAP), accuracy, and diagnostic time. Results: The best-performing model achieved 89.31% recall, 84.23% precision, an F1 score of 0.8670, and an mAP of 0.817 on the validation set. On the independent test set, the model achieved 93.89% accuracy. In addition, the model required substantially less time for image interpretation than the participating clinicians. Conclusion: The proposed YOLOv6-based framework showed good performance for bacterial identification in sputum M-ROSE smears under limited annotated data conditions. These findings support the feasibility of applying data-efficient deep learning strategies to real-world clinical microbiological images and suggest potential utility in rapid microbiological diagnosis.

Objectives Adult-onset Still’s disease (AOSD)-associated macrophage activation syndrome (MAS) is a highly inflammatory condition. Currently, the main treatment method for AOSD-associated MAS is high-dose glucocorticoid shock therapy, but the effect is not satisfactory, and new therapeutic methods or drugs are urgently needed. The aim of this study was to evaluate the efficacy and safety of salvage baricitinib therapy for AOSD-associated MAS following conventional immunosuppressive therapy. Method Data from nine patients with MAS treated in the Department of Infection and the Department of Rheumatology, First Affiliated Hospital of Soochow University, from January 2020 to December 2022, were retrospective analysed. Salvage therapy with baricitinib was administered, with an initial dose of 8 mg/day. The patients’ clinical symptoms and MAS indices (blood counts and triglyceride, fibrinogen, lactate dehydrogenase, sCD25, and Fer levels) were observed. After a complete response was achieved, the baricitinib dose was gradually reduced to 2–4 mg/day and maintained. Results The MAS blood test results of all nine patients improved to varying degrees within 1 week post-treatment. The MAS indices gradually returned to normal approximately 8 weeks after discharge; at the 12-month outpatient follow-up, seven patients did not develop MAS again, but two patients (patients 5 and 9) had MAS recurrence. The follow-up period ended with salvage treatment with emapalumab. Conclusions Baricitinib can effectively inhibit the inflammatory response and improve outcomes in the treatment of MAS secondary to rheumatic immunity-related diseases; moreover, this drug treatment is safe. Our study provides a new strategy for MAS salvage therapy. Key Points • Bariticinib may be a palliative treatment option for recurrent/refractory MAS.

Background and Aims: To determine appropriate dosages for mice models in liver failure prophase and explore IL-35 and cytokine levels. Also, investigate changes in Treg/ Th17 ratio, IL-35, and IL-17 in patients with HBV-induced liver failure before and after treatment. Methods: We induced liver failure in mice with different doses of concanavalin and assessed severity through serum biochemistry and pathology. Mice were divided into control, model, IL-35 plasmid, anti-IL-35, and dexamethasone groups. IL-35 and IL-17 expressions in liver tissue were evaluated via immunohistochemistry, fluorescence staining, and ELISA. Treg/Th17 ratio and IL-35 and IL-17 levels in peripheral blood of patients were measured using flow cytometry and ELISA, compared to 22 healthy individuals. Results: The optimal dose for the liver failure model was 20 μ g/ kg concanavalin. The IL-35 plasmid and dexamethasone groups showed significantly lower serum TBil, ALT, AST, IL-4, IL-17, TNF-α , and liver histopathology scores compared to the model group, while the anti-IL-35 group showed higher levels (P<0. 05). In HBV-PLF patients, TBil, ALT, and AST significantly decreased, and PTA increased after glucocorticoid treatment (P<0. 05). Treg/ Th17 ratio was lower in HBV-PLF patients compared to the healthy group (P<0. 05), with higher IL-35 and IL-17 levels (P<0. 05). Post-treatment, Treg increased and Th17 decreased significantly; IL-17 and Treg/ Th17 ratios increased, while IL-35 decreased (P<0.05). IL-35 positively correlated with the Treg/Th17 ratio. Conclusions 20 μ g/ kg concanavalin proper dose for mice model. IL-35 may protect in liver failure. Glucocorticoids help maintain immune balance, prevent failure, increase ratios and IL-35 in HBV-PLF patients, and IL-35 related to Treg/Th17 balance.

Background: Acute liver injury (ALI) is characterized by massive hepatocyte death and has high mortality and poor prognosis. Hepatocyte pyroptosis plays a key role in the pathophysiology of ALI and is involved in the inflammatory response mediated by NOD-like receptor protein 3 (NLRP3) inflammasome activation. Deltex 1 (DTX1) is a single transmembrane protein with ubiquitin E3 ligase activity and is closely involved in cell growth, differentiation, and apoptosis, as well as intracellular signal transduction. However, little is known about the influence of DTX1 on ALI. This study aimed to investigate the role of DTX1 in pyroptosis and inflammation induced by D-galactosamine (D-GalN) and tumor necrosis factoralpha (TNF-α) in human hepatocytes (LO2 cells) in vitro. Methods: Cell pyroptosis was measured by flow cytometry. The levels of DTX1, pyroptosis-associated proteins, and inflammatory cytokines were detected by quantitative real-time polymerase chain reaction, western blotting, and enzyme-linked immunosorbent assay. Immunofluorescence staining, co-immunoprecipitation, ubiquitination, and luciferase reporter and chromatin immunoprecipitation assays were performed to detect the regulation between DTX1 and NLRP3 or hepatocyte nuclear factor 4 alpha (HNF4α). Analysis of variance was performed to compare groups. Results: We found that DTX1 was decreased in D-GalN/TNF-α-induced LO2 cells. DTX1 overexpression significantly inhibited D-GalN/TNF-α-induced cell pyroptosis and inflammation. DTX1 interacted with NLRP3 and induced NLRP3 ubiquitination and degradation. Furthermore, by targeting NLRP3, DTX1 knockdown significantly induced cell pyroptosis and inflammation. In addition, HNF4α promoted DTX1 transcription by binding with its promoter. Conclusion: Our study revealed that DTX1 suppressed D-GalN/TNF-α-induced hepatocyte pyroptosis and inflammation by regulating NLRP3 ubiquitination.

Epstein–Barr virus-associated hemophagocytic lymphohistiocytosis (EBV-HLH) is a severe hyperinflammatory illness that affects adults and is caused by an EBV infection. Without allogeneic hematopoietic stem cell transplantation (allo-HSCT), the overall survival of adult patients with EBV-HLH is unsatisfactory, necessitating the development of innovative therapeutic approaches. The clinical records of twelve EBV-HLH patients who received sintilimab therapy combined with ruxolitinib on a compassionate basis at the First Affiliated Hospital of Soochow University were retrospectively examined in this investigation. All the patients responded without fever, but three patients relapsed within a week. Among the nine patients achieving complete response (CR), 55.6% (5/9) maintained CR for >4.5 months, and 33.3% (3/9) relapsed following CR. Neither patients with no response (NR) nor relapsed patients were fit for allo-HSCT, and all died soon after discharge. Six patients had clinical CR with a median follow-up of 5 (4.4–14.7) months. There were no documented severe negative effects. Additional information on this innovative treatment for adult EBV-HLH is provided in our report.

Sputum smear tests are critical for the diagnosis of respiratory diseases. Automatic segmentation of bacteria from sputum smear images is important for improving diagnostic efficiency. However, this remains a challenging task owing to the high interclass similarity among different categories of bacteria and the low contrast of the bacterial edges. To explore more levels of global pattern features to promote the distinguishing ability of bacterial categories and maintain sufficient local fine-grained features to ensure accurate localization of ambiguous bacteria simultaneously, we propose a novel dual-branch deformable cross-attention fusion network (DB-DCAFN) for accurate bacterial segmentation. Specifically, we first designed a dual-branch encoder consisting of multiple convolution and transformer blocks in parallel to simultaneously extract multilevel local and global features. We then designed a sparse and deformable cross-attention module to capture the semantic dependencies between local and global features, which can bridge the semantic gap and fuse features effectively. Furthermore, we designed a feature assignment fusion module to enhance meaningful features using an adaptive feature weighting strategy to obtain more accurate segmentation. We conducted extensive experiments to evaluate the effectiveness of DB-DCAFN on a clinical dataset comprising three bacterial categories: Acinetobacter baumannii, Klebsiella pneumoniae, and Pseudomonas aeruginosa. The experimental results demonstrate that the proposed DB-DCAFN outperforms other state-of-the-art methods and is effective at segmenting bacteria from sputum smear images.

Objective: To study the construction of a prognostic model for hepatocellular carcinoma (HCC) based on pyroptosis-related genes (PRGs). Methods: HCC patient datasets were obtained from the Cancer Genome Atlas (TCGA) database, and a prognostic model was constructed by applying univariate Cox and least absolute shrinkages and selection operator (LASSO) regression analysis. According to the median risk score, HCC patients in the TCGA dataset were divided into high-risk and low-risk groups. Kaplan-Meier survival analysis, receiver operating characteristic (ROC) curves, univariate and multivariate Cox analysis, and nomograms were used to evaluate the predictive ability of the prognostic models. Functional enrichment analysis and immune infiltration analysis were performed on differentially expressed genes between the two groups. Finally, two HCC datasets (GSE76427 and GSE54236) from the Gene Expression Omnibus database were used to externally validate the prognostic value of the model. Univariate and multivariate Cox regression analysis or Wilcoxon tests were performed on the data. Results: A total of 366 HCC patients were included after screening the HCC patient dataset obtained from the TCGA database. A prognostic model related to HCC was established using univariate Cox regression analysis, LASSO regression analysis, and seven genes (CASP8, GPX4, GSDME, NLRC4, NLRP6, NOD2, and SCAF11). 366 cases were evenly divided into high-risk and low-risk groups based on the median risk score. Kaplan-Meier survival analysis showed that there were statistically significant differences in the survival time between patients in the high-risk and low-risk groups in the TCGA, GSE76427, and GSE54236 datasets (median overall survival time was 1 149 d vs. 2 131 d, 4.8 years vs. 6.3 years, and 20 months vs. 28 months, with P = 0.000 8, 0.034 0, and 0.0018, respectively). ROC curves showed good survival predictive value in both the TCGA dataset and two externally validated datasets. The areas under the ROC curves of 1, 2, and 3 years were 0.719, 0.65, and 0.657, respectively. Multivariate Cox regression analysis showed that the risk score of the prognostic model was an independent predictor of overall survival time in HCC patients. The risk model score accurately predicted the survival probability of HCC patients according to the established nomogram. Functional enrichment analysis and immune infiltration analysis showed that the immune status of the high-risk group was significantly decreased. Conclusion: The prognostic model constructed in this study based on seven PRGs accurately predicts the prognosis of HCC patients.

Luteolin is a natural flavonoid that has a variety of pharmacological activities, such as anti-inflammatory, anti-allergic, anti-bacterial, anti-viral, apoptosis inhibition, cell autophagy regulation, and anti-tumor activity. It is one of the main ingredients of an expert-recommended herbal formula for the prevention and treatment of coronavirus disease 2019 (COVID-19). This suggests that luteolin has strong pharmacological effects on the prevention and treatment of COVID-19. The aims of this study were to identify the molecular targets of luteolin and to infer the possible mechanisms by which it exerts its pharmacological effects. The GSE159787 data set was obtained from the Gene Expression Omnibus online database, and differentially expressed genes were analyzed. There were 22 upregulated differentially expressed genes enriched in the COVID-19 signaling pathway, suggesting that the upregulation of these genes may be closely related to the occurrence of COVID-19. Molecular docking results showed that luteolin had strong binding efficiency to 20 of these 22 key genes. Six of these genes (CFB, EIF2AK2, OAS1, MAPK11, OAS3, and STAT1) showed strong binding activity. Luteolin can regulate the COVID-19 signaling pathway by combining with these targets, which may have a therapeutic effect on COVID-19.

Human adenovirus (HAdV) is a common virus, but the infections it causes are relatively uncommon. At the same time, the methods for the detection of HAdV are varied, among which viral culture is still the gold standard. HAdV infection is usually self-limited but can also cause clinically symptomatic in lots of organs and tissues, of which human adenovirus pneumonia is the most common. In contrast, human adenovirus hepatitis is rarely reported. However, HAdV hepatitis has a high fatality rate once it occurs, especially in immunocompromised patients. Although human adenovirus hepatitis has some pathological and imaging features, its clinical symptoms are not typical. Therefore, HAdV hepatitis is not easy to be found in the clinic. There are kinds of treatments to treat this disease, but few are absolutely effective. In view of the above reasons, HAdV hepatitis is a disease that is difficult to be found in time. We reviewed and summarized the previously reported cases, hoping to bring some relatively common characteristics to clinicians, so as to facilitate early detection, early diagnosis, and early treatment of patients.