Rethinking Scientific Summarization Evaluation: Grounding Explainable Metrics on Facet-aware Benchmark.

The summarization capabilities of pretrained and large language models (LLMs) have been widely validated in general areas, but their use in scientific corpus, which involves complex sentences and specialized knowledge, has been less assessed. This paper presents conceptual and experimental analyses of scientific summarization, highlighting the inadequacies of traditional evaluation methods, such as n-gram, embedding comparison, and QA, particularly in providing explanations, grasping scientific concepts, or identifying key content. Subsequently, we introduce the Facet-aware Metric (FM), employing LLMs for advanced semantic matching to evaluate summaries based on different aspects. This facet-aware approach offers a thorough evaluation of abstracts by decomposing the evaluation task into simpler subtasks. Recognizing the absence of an evaluation benchmark in this domain, we curate a Facet-based scientific summarization Dataset (FD) with facet-level annotations. Our findings confirm that FM offers a more logical approach to evaluating scientific summaries. In addition, fine-tuned smaller models can compete with LLMs in scientific contexts, while LLMs have limitations in learning from in-context information in scientific domains. This suggests an area for future enhancement of LLMs.

Dual-responsive nanoplatform for integrated cancer diagnosis and therapy: Unleashing the power of tumor microenvironment.

Chemodynamic therapy (CDT), designed to trigger a tumor-specific hydrogen peroxide (H2O2) reaction generating highly toxic hydroxyl radicals (·OH), has been investigated for cancer treatment. Unfortunately, the limited Fenton or Fenton-like reaction rate and the significant impact of excessive reducing glutathione (GSH) in the tumor microenvironment (TME) have severely compromised the effectiveness of CDT. To address this issue, we designed a dual-responsive nanoplatform utilizing a metal-polyphenol network (MPN) -coated multi-caged IrOx for efficient anti-tumor therapy in response to the acidic TME and intracellular excess of GSH, in which MPN composed of Fe3+ and tannic acid (TA). Initially, the acidic TME and intracellular excess of GSH lead to the degradation of the MPN shell, resulting in the release of Fe3+ and exposure of the IrOx core, facilitating the efficient dual-pathway CDT. Subsequently, the nanoplatform can mitigate the attenuation of CDT by consuming the excessive GSH within the tumor. Finally, the multi-caged structure of IrOx is advantageous for effectively implementing photothermal therapy (PTT) in coordination with CDT, further enhancing the therapeutic efficacy of tumors. Moreover, the outstanding Computed Tomography (CT) and Magnetic Resonance Imaging (MRI) (T1/T2) multimodal imaging capabilities of IrOx@MPN enable early diagnosis and timely treatment. This work provides a typical example of the construction of a novel multifunctional platform for dual-responsive treatment of tumors.

An AI Agent for Fully Automated Multi-Omic Analyses.

With the fast-growing and evolving omics data, the demand for streamlined and adaptable tools to handle bioinformatics analysis continues to grow. In response to this need, Automated Bioinformatics Analysis (AutoBA) is introduced, an autonomous AI agent designed explicitly for fully automated multi-omic analyses based on large language models (LLMs). AutoBA simplifies the analytical process by requiring minimal user input while delivering detailed step-by-step plans for various bioinformatics tasks. AutoBA's unique capacity to self-design analysis processes based on input data variations further underscores its versatility. Compared with online bioinformatic services, AutoBA offers multiple LLM backends, with options for both online and local usage, prioritizing data security and user privacy. In comparison to ChatGPT and open-source LLMs, an automated code repair (ACR) mechanism in AutoBA is designed to improve its stability in automated end-to-end bioinformatics analysis tasks. Moreover, different from the predefined pipeline, AutoBA has adaptability in sync with emerging bioinformatics tools. Overall, AutoBA represents an advanced and convenient tool, offering robustness and adaptability for conventional multi-omic analyses.

Clinical characteristics and nomogram model for predicting the risk of recurrence of complicated urinary tract infection in pediatric patients.

Complicated urinary tract infection (cUTI) has higher incidences of antibiotic resistance, recurrence, chronicity, and progression. There exist differences in the immune functions, anatomical structures and physiological functions of urinary tract system between children and adults. In addition, the immature stage of hygiene habit development and the inability to express discomfort accurately all contribute to cUTI recurrence in pediatric patients. cUTI recurrence in children can lead to serious consequences such as growth and development delay, renal scars, and progression to end-stage kidney diseases. Despite the high incidence of cUTI recurrence, no predictive model currently exists to guide targeted intervention in pediatric patients. This study aimed to establish a nomogram to predict the risk of cUTI recurrence in children, thereby facilitating targeted prevention and treatment strategies in pediatric patients. The nomogram was developed based on a retrospective cohort that included 421 pediatric patients with cUTI at West China Second University Hospital from January 2020 to August 2023. The patients were randomly divided into a training set and a validation set in a 3:1 ratio. Logistic regression analysis was used to identify risk factors and construct the nomogram for predicting the risk of cUTI recurrence, followed by validation and performance analysis. Of the 421 children with cUTI, the recurrence rate of cUTI was 68.4% (288 cases) during an average follow-up duration of 22.9 months. The nomogram comprised female gender, history of urinary tract surgery, Escherichia coli in urine culture, renal dysfunction, and vesicoureteral reflux as predictors of cUTI recurrence in pediatric patients. The model showed favorable performance with C-index values of 0.735 and 0.750 in the training dataset and the validation dataset, respectively. The decision curve analysis revealed that the nomogram might be clinically useful. The reliable nomogram would be beneficial for clinicians to identify children with high risks of cUTI recurrence for targeted intervention.

Perinatal outcomes with a copper intrauterine device in situ after 28 gestational weeks: A retrospective study of sole center in southeast China.

The purpose of this study was to investigate the pregnancy outcomes with a copper intrauterine device (IUD) in situ after 28 gestational weeks and the association between pregnancy with copper IUDs and neonatal congenital malformations. This retrospective study had compared the singleton pregnancies with the copper IUDs in situ and without after 28 gestational weeks in 1 delivery center of southeast China. The main exposure was a copper IUD in uterine cavity with pregnancy. The pregnant outcomes as preterm birth, premature rupture of membranes, infections were observed and compared. We had also compared neonatal congenital malformations in 2 groups. The statistical analysis was carried out using R (version 4.0.4; R Development Core Team) statistical software. Association between IUD use or duration of IUD use and adverse pregnancy outcomes were estimated using logistic model. Two-tailed P value < .05 was deemed statistically significant. A total of 148 pregnant women were included in our study, 74 with copper IUDs in situ were categorized into case group and 74 without IUDs during pregnancy into control group. No significant difference of maternal age, BMI, birth weight and gender were observed between 2 groups. In case group, the rates of preterm premature rupture of membranes (37.8%) and spontaneous preterm birth (23.0%) were significantly high compared to control group. Odds ratios of premature rupture of membranes and spontaneous preterm birth were 2.86 and 5.22 respectively. Women of elder age (≥35 years) in case group were more likely to experience premature rupture of membranes. The rates of neonatal congenital malformation were 10.8% (8/74) in case group and 1.4% (1/74) in control group respectively. We had found that pregnancy with copper IUD in situ increased the risk of premature rupture of membranes and spontaneous preterm birth after 28 gestation weeks, the risk of spontaneous preterm birth increased 5.22 times. Pregnancy with IUD in situ may be at increased risk of infection and neonatal malformation.

Convenient analysis of sartan adulteration in herbal oral liquids using cotton fiber-supported liquid extraction and with high-performance liquid chromatography-fluorescence detection.

This research introduces a novel approach for detecting sartan antihypertensive drug adulteration in herbal oral liquids using cotton fiber-supported liquid extraction (CF-SLE) combined with high-performance liquid chromatography-fluorescence detection (HPLC-FLD). Optimal extraction parameters were determined through systematic method development, establishing a sample solution with a pH of 3.0, using 200 mg of cotton fiber, ethyl acetate as the extraction solvent, and a solvent volume of 4 mL. These conditions demonstrated robust extraction efficiency and were further validated for precision and accuracy, with intra- and inter-day relative standard deviations consistently below 7.5 % and relative recoveries ranging from 88.5 % to 106.1 %. The method exhibited excellent linearity for sartans, with R² values greater than 0.993 across a concentration range of 10-2000 ng/mL. Detection limits were effectively established in the range of 2.6-3.1 ng/mL, indicating that the method's sensitivity is adequate for the intended screening purposes. This validated method was then applied to real sample analysis, confirming its potential for routine use in detecting illegal additives within complex herbal matrices, thereby ensuring consumer safety and supporting regulatory compliance.

Single-cell RNA Sequencing Identifies a Novel Subtype of Microglia with High Cd74 Expression that Facilitates White Matter Inflammation During Chronic Cerebral Hypoperfusion.

Vascular dementia (VaD) causes progressive cognitive decline in the elderly population, but there is short of available therapeutic measures. Microglia-mediated neuroinflammation is vigorously involved in the pathogenesis of VaD, but the traditional classification of microglial M1/M2 phenotypes remains restrictive and controversial. This study aims to investigate whether microglia transform into novel subtypes in VaD. Chronic cerebral hypoperfusion (CCH) rat model was constructed to mimic VaD. Microglia were isolated via magnetic-activated cell sorting and analyzed by single-cell RNA sequencing (scRNA-seq) and bioinformatics. The findings inferred from scRNA-seq and bioinformatics were further validated through in vivo experiments. In this study, microglia were divided into eight clusters. The proportion of MG5 cluster was significantly increased in the white matter of the CCH group compared with the Sham group and was named chronic ischemia-associated microglia (CIAM). Immunity- and inflammation-related genes, including RT1-Db1, RT1-Da, RT1-Ba, Cd74, Spp1, C3, and Cd68, were markedly upregulated in CIAM. Enrichment analysis illustrated that CIAM possessed the function of evoking neuroinflammation. Further studies unveiled that Cd74 is associated with the most abundant GO terms involved in inflammation as well as cell proliferation and differentiation. In addition, microglia-specific Cd74 knockdown mediated by adeno-associated virus decreased the abundance of CIAM in the white matter, thereby mitigating inflammatory cytokine levels, alleviating white matter lesions, and improving cognitive impairment for CCH rats. These findings indicate that Cd74 is the core molecule of CIAM to trigger neuroinflammation and induce microglial differentiation to CIAM, suggesting that Cd74 may be a potential therapeutic target for VaD.

Pre-trained multimodal large language model enhances dermatological diagnosis using SkinGPT-4.

Large language models (LLMs) are seen to have tremendous potential in advancing medical diagnosis recently, particularly in dermatological diagnosis, which is a very important task as skin and subcutaneous diseases rank high among the leading contributors to the global burden of nonfatal diseases. Here we present SkinGPT-4, which is an interactive dermatology diagnostic system based on multimodal large language models. We have aligned a pre-trained vision transformer with an LLM named Llama-2-13b-chat by collecting an extensive collection of skin disease images (comprising 52,929 publicly available and proprietary images) along with clinical concepts and doctors' notes, and designing a two-step training strategy. We have quantitatively evaluated SkinGPT-4 on 150 real-life cases with board-certified dermatologists. With SkinGPT-4, users could upload their own skin photos for diagnosis, and the system could autonomously evaluate the images, identify the characteristics and categories of the skin conditions, perform in-depth analysis, and provide interactive treatment recommendations.

Leveraging Professional Radiologists' Expertise to Enhance LLMs' Evaluation for Radiology Reports.

In radiology, Artificial Intelligence (AI) has significantly advanced report generation, but automatic evaluation of these AI-produced reports remains challenging. Current metrics, such as Conventional Natural Language Generation (NLG) and Clinical Efficacy (CE), often fall short in capturing the semantic intricacies of clinical contexts or overemphasize clinical details, undermining report clarity. To overcome these issues, our proposed method synergizes the expertise of professional radiologists with Large Language Models (LLMs), like GPT-3.5 and GPT-4. Utilizing In-Context Instruction Learning (ICIL) and Chain of Thought (CoT) reasoning, our approach aligns LLM evaluations with radiologist standards, enabling detailed comparisons between human and AI-generated reports. This is further enhanced by a Regression model that aggregates sentence evaluation scores. Experimental results show that our "Detailed GPT-4 (5-shot)" model achieves a 0.48 score, outperforming the METEOR metric by 0.19, while our "Regressed GPT-4" model shows even greater alignment with expert evaluations, exceeding the best existing metric by a 0.35 margin. Moreover, the robustness of our explanations has been validated through a thorough iterative strategy. We plan to publicly release annotations from radiology experts, setting a new standard for accuracy in future assessments. This underscores the potential of our approach in enhancing the quality assessment of AI-driven medical reports.

Metabolic disorder: the dark side of ovarian aging.

Ovarian aging plays an important role in the aging process of the whole body. It has been reported that metabolic disorder may significantly contribute to ovarian aging. This article highlights recent advances in metabolic regulation of ovarian aging and highlights key issues in the field.

TET1 inhibits the migration and invasion of cervical cancer cells by regulating autophagy.

Methylation modifications play pertinent roles in regulating gene expression and various biological processes. The silencing of the demethylase enzyme TET1 can affect the expressions of key oncogenes or tumour suppressor genes, thus contributing to tumour formation. Nonetheless, how TET1 affects the progression of cervical cancer is yet to be elucidated. In this study, we found that the expression of TET1 was significantly downregulated in cervical cancer tissues. Functionally, TET1 knockdown in cervical cancer cells can promote cell proliferation, migration, invasion, cervical xenograft tumour formation and EMT. On the contrary, its overexpression can reverse the aforementioned processes. Moreover, the autophagy level of cervical cancer cells can be enhanced after TET1 knockdown. Mechanistically, methylated DNA immunoprecipitation (MeDIP)-sequencing and MeDIP quantitative real-time PCR revealed that TET1 mediates the methylation of autophagy promoter regions. These findings suggest that TET1 affects the autophagy of cervical cancer cells by altering the methylation levels of NKRF or HIST1H2AK, but the specific mechanism needs to be investigated further.

Is early bilateral compression ultrasonography and D-dimer monitoring appropriately for prophylaxis and diagnosis of deep venous thrombosis after cesarean section women: a single-center observation study of Chinese Han population.

BACKGROUND: Venous thromboembolism (VTE) is most prevalent among parturients following a cesarean section (CS). The objective of this study was to assess the practical utility of bilateral compression ultrasonography (CUS) of the lower limbs, coupled with D-dimer monitoring, in the early diagnosis of VTE within the Han Chinese population. METHODS: Our prospective observational study included 742 women who underwent CUS and D-dimer testing on the first day post-CS. Subsequently, telephone or outpatient follow-ups were conducted until 42 days postpartum. States of hypercoagulation and thrombosis, as indicated by CUS, were classified as CUS abnormal. A D-dimer level ≥ 3 mg/l was considered the D-dimer warning value. Early ambulation and mechanical prophylaxis were universally recommended for all parturients post-CS. A sequential diagnostic strategy, based on the 2015 RCOG VTE risk-assessment tool, was employed. Therapeutic doses of low-molecular-weight heparin (LMWH) were administered for the treatment of thromboembolic disease. Prophylactic doses of LMWH were given for VTE prophylaxis in parturients with hypercoagulative status accompanied by D-dimer levels ≥ 3 mg/l. All high-risk women (RCOG score ≥ 4 points) were additionally treated with preventive LMWH. Statistical analyses were conducted using the R statistical software, with a two-sided P value < 0.05 considered statistically significant. RESULTS: Fifteen cases of VTE and 727 instances without VTE were observed. The overall VTE rate post-CS was 2.02% (15/742), with 66.7% (10/15) being asymptomatic. Eleven patients received a VTE diagnosis on the first postpartum day. Among the 41 parturients exhibiting hypercoagulation ultrasound findings and D-dimer levels ≥ 3 mg/l, despite receiving pharmacological VTE prophylaxis with LMWH, 4.88% (2/41) in the high-risk group were eventually diagnosed with VTE. A total of 30.86% (229/742) exhibited normal ultrasound findings and D-dimer levels < 3 mg/l on the first day post-CS, with no VTE occurrences in the postpartum follow-up. According to RCOG's recommendation, 78.03% (579/742) of cesarean delivery women should receive prophylactic anticoagulation, while only 20.62% (153/742) met our criterion for prophylactic anticoagulation. CONCLUSION: The strategy of timely routine bilateral CUS and D-dimer monitoring is conducive to the early diagnosis and treatment of VTE, significantly reducing the use of LMWH in the Chinese Han population.

Afterdischarges in myotonic dystrophy type 1.

OBJECTIVE: Electrodiagnostic testing is an important screening test for myotonic dystrophy type 1 (DM1). Although myotonic discharges are observed on electromyography in cases of DM1, it is difficult to distinguish DM1 from other myotonic disorders clinically. In the present study, afterdischarges, another type of pathological potential revealed by electrodiagnostic testing, were analyzed, and their role in distinguishing DM1 from other myotonic disorders was explored. METHODS: Data from 33 patients with myotonic discharges on electromyography were analyzed retrospectively. According to gene testing, the patients were divided into DM1 (n = 20) and non-DM1 myotonia (n = 13) groups. Afterdischarges were investigated by retrospectively evaluating the electrodiagnostic findings of motor nerve conduction studies, F-waves, and repetitive nerve stimulations. RESULTS: Afterdischarges were observed in 17 of the 20 patients with DM1, with an occurrence rate of approximately 85%. However, afterdischarges were absent in all patients with non-DM1 myotonia. There were significant differences in the occurrence rate between the two groups (P < 0.01). CONCLUSION: Afterdischarges may serve as a suggestive role in clinical diagnosis of DM1. The discovery that DM1 can present with afterdischarges may pave a new way to study the pathogenesis of DM1.

Impact of lamivudine treatment in late pregnancy on the development of the foetal immune response to hepatitis B virus: a meta-analysis in R with the metafor package.

BACKGROUND: Hepatitis B virus (HBV) infection is a worldwide public health burden, especially in Asia and Africa. Concerns were raised that foetal exposure to HBV and antiretroviral therapy (ART) might suppress the innate immune response and reduce the production of hepatitis B surface antibody (HBsAb) in foetuses and infants. We therefore conducted the current study to evaluate the impact of ART on the development of the immune response to HBV in foetuses and infants. METHODS: We selected lamivudine instead of telbivudine or tenofovir as the intervention measurement because it was the oldest and most widely used ART during pregnancy and its safety data have been sufficiently documented. A comprehensive search was conducted in eight electronic databases, including four Chinese and four English databases. Studies that met the following eligibility criteria were included: human randomized controlled trials (RCTs); participants in the treatment group were exclusively exposed to lamivudine; participants in the control group were exposed to placebo, no treatment or hepatitis B immunoglobulin; all participants were HBV-positive pregnant women with a high viral load and the main outcome of interest was neonatal HBsAb seropositivity. Data were tabulated and analysed using R software. RESULTS: Nine RCTs were included and analysed. Compared with controls, lamivudine significantly decreased HBsAb seronegativity in the newborn within 24 h after birth (indicating the foetal immune response to HBV). Similar results were noted in infants within 6-7 months after birth and infants within 12 months (indicating the neonatal immune response to HBV vaccine). CONCLUSIONS: Lamivudine treatment in late pregnancy boosted the foetal immune response to HBV in utero and enhanced the neonatal immune response to hepatitis B vaccine after birth.

Serum Proteomics Identified TAFI as a Potential Molecule Facilitating the Migration of Peripheral Monocytes to Damaged White Matter During Chronic Cerebral Hypoperfusion.

Neuroinflammation is assumed as the critical pathophysiologic mechanism of white matter lesions (WMLs), and infiltrated peripheral monocyte-derived macrophages are implicated in the development of neuroinflammation. This study sought to explore the blood molecules that promote the migration of peripheral monocytes to the sites of WMLs. The serum protein expression profiles of patients and Sprague-Dawley rat models with WMLs were detected by data-independent acquisition (DIA) proteomics technique. Compared with corresponding control groups, we acquired 62 and 41 differentially expressed proteins (DEPs) in the serum of patients and model rats with WMLs respectively. Bioinformatics investigations demonstrated that these DEPs were linked to various Kyoto Encyclopedia of Genes and Genomes (KEGG) pathways and Gene Ontology (GO) terms involved in neuroinflammation. Afterward, we identified thrombin-activatable fibrinolysis inhibitor (TAFI) as a shared and overexpressed protein in clinical and animal serum samples, which was further verified by enzyme-linked immunosorbent assay. Additionally, an upregulation of TAFI was also observed in the white matter of rat models, and the inhibition of TAFI impeded the migration of peripheral monocytes to the area of WMLs. In vitro experiments suggested that TAFI could enhance the migration ability of RAW264.7 cells and increase the expression of Ccr2. Our study demonstrates that neuroinflammatory signals can be detected in the peripheral blood of WMLs patients and model rats. TAFI may serve as a potential protein that promotes the migration of peripheral monocytes to WMLs regions, thereby providing a novel molecular target for further investigation into the interaction between the central and peripheral immune systems.

The Impact of Dipyridamole on Disease-Associated Microglia Phenotypic Transformation in White Matter Lesions Induced by Chronic Cerebral Hypoperfusion.

White matter lesions (WMLs) resulting from chronic cerebral hypoperfusion (CCH) are the leading cause of vascular dementia (VaD). This study aimed to investigate whether dipyridamole could alleviate WMLs by regulating the phenotype of disease-associated microglia (DAM) through equilibrative nucleoside transporter 2 (ENT2) and adenosine A2A receptor (Adora2a) and to clarify the underlying molecular mechanisms. CCH rat models were constructed to mimic VaD. Morris water maze and Luxol Fast Blue staining were employed to assess cognitive function and quantify the severity of WMLs, respectively. Immunofluorescent staining was performed to analyze the activation of glial cells and the phenotypic transformation of DAM. Additionally, levels of ENT2, proteins in the NF-κB and ERK1/2 pathways and inflammatory cytokines were detected. The results indicated that dipyridamole diminished the activation and proliferation of microglia and astrocytes, increased the expression of myelin basic protein and ameliorated WMLs and cognitive decline in CCH rats. Further study revealed that dipyridamole decreased the expression of ENT2 and inhibited the activation of ERK1/2 and NF-κB signaling pathways, which ultimately converted DAM to anti-inflammatory phenotype and suppressed the levels of TNF-α, IL-1ß, IL-6 in WMLs. However, Adora2a inhibitor (SCH58261) attenuated above effects. Our study demonstrates that dipyridamole facilitates the conversion of DAM to the anti-inflammatory phenotype through ENT2/Adora2a pathway and inhibits the activation of ERK1/2 and NF-κB signaling pathways, thereby alleviating neuroinflammation in WMLs. The current findings establish the basis for using dipyridamole to treat VaD.

Development and validation of models for predicting preterm birth and gestational latency following emergency cervical cerclage: A multicenter cohort study.

INTRODUCTION: Emergency cervical cerclage is a recognized method for preventing mid-trimester pregnancy loss and premature birth; however, its benefits remain controversial. This study aimed to establish preoperative models predicting preterm birth and gestational latency following emergency cervical cerclage in singleton pregnant patients with a high risk of preterm birth. MATERIAL AND METHODS: We retrospectively reviewed data from patients who received emergency cerclage between 2015 and 2023 in three institutions. Patients were grouped into a derivation cohort (n = 141) and an independent validation cohort (n = 61). Univariate and multivariate logistic and Cox regression analyses were used to identify independent predictive variables and establish the models. Harrell's C-index, time-dependent receiver operating characteristic curves and areas under the curves, calibration curve, and decision curve analyses were performed to assess the models. RESULTS: The models incorporated gestational weeks at cerclage placement, history of prior second-trimester loss and/or preterm birth, cervical dilation, and preoperative C-reactive protein level. The C-index of the model for predicting preterm birth before 28 weeks was 0.87 (95% CI: 0.82-0.93) in the derivation cohort and 0.82 (95% CI: 0.71-0.92) in the independent validation cohort; The C-index of the model for predicting gestational latency was 0.70 (95% CI: 0.66-0.75) and 0.78 (95% CI: 0.71-0.84), respectively. In the derivation set, the areas under the curves were 0.84, 0.81, and 0.84 for predicting 1-, 3- and 5-week pregnancy prolongation, respectively. The corresponding values for the external validation were 0.78, 0.78, and 0.79, respectively. Calibration curves showed a good homogeneity between the observed and predicted ongoing pregnant probabilities. Decision curve analyses revealed satisfactory clinical usefulness. CONCLUSIONS: These novel models provide reliable and valuable prognostic predictions for patients undergoing emergency cerclage. The models can assist clinicians and patients in making personalized clinical decisions before opting for the cervical cerclage.

Varicella-associated disseminated intravascular coagulation secondary to Henoch-Schönlein purpura with renal and gastrointestinal system involvement in a child: A case report.

RATIONALE: Immunocompromised patients who developed varicella-zoster virus (VZV)-associated disseminated intravascular coagulation (DIC) previously included recipients of bone marrow, hematopoietic stem cell, or organ transplantations, patients with primary nephropathy receiving corticosteroid therapy, cancer patients receiving chemotherapy, and patients with human immune deficiency virus infection. The case reported here is novel because, to our knowledge, there has been no report of VZV-associated DIC after the onset of Henoch-Schönlein purpura (HSP). PURPOSE: To report the successful treatment of a novel pediatric case with VZV-associated DIC secondary to HSP. DIAGNOSIS AND INTERVENTION: An 8-year-old girl developed VZV-associated DIC 24 days after diagnosis of HSP with renal and gastrointestinal involvement. She was treated with methylprednisolone at a local hospital for 19 days, and suddenly developed fever starting from day 4 in our hospital. Her fever persisted with vesicular skin rashes on her back, strong abdominal and lower back pain, epistaxis, hematochezia, erosion and bleeding on her lips, in her mouth and at puncture sites on day 5. She was diagnosed with DIC with the laboratory evidence of dramatically decreased platelet count and fibrinogen, prolonged activated partial thromboplastin time and prothrombin time, and increased fibrin degradation products including d-dimers. She also developed multiple organ dysfunction syndrome. On day 7, the patient VZV nucleic acid result turned out to be positive. Methylprednisolone treatment was discontinued, and she was given a multi-modality therapy including medications of acyclovir and antibiotics, intravenous gamma-immunoglobulin, various blood product transfusions, continuous renal replacement therapy, plasma exchange, and administration of liver and gastrointestinal system protection drugs. OUTCOMES: The patient multi-organ function damage gradually recovered. After VZV control, the patient was treated with oral methylprednisolone again for HSP with nephritis. Urine analysis was normal 1 year later, and oral hormone was discontinued. No complication or relapse occurred during 2 years of follow-up. SIGNIFICANCE: This case report, for the first time, adds HSP treated with corticosteroids to the spectrum of clinical conditions that progressed to life-threatening secondary varicella-associated DIC. Early identification of varicella infection and DIC, combined with timely antiviral, immunoglobulin transfusion, plasma exchange, and other combined therapies are essential for saving patients' lives.

Opportunities and Challenges for ChatGPT and Large Language Models in Biomedicine and Health.

ChatGPT has drawn considerable attention from both the general public and domain experts with its remarkable text generation capabilities. This has subsequently led to the emergence of diverse applications in the field of biomedicine and health. In this work, we examine the diverse applications of large language models (LLMs), such as ChatGPT, in biomedicine and health. Specifically we explore the areas of biomedical information retrieval, question answering, medical text summarization, information extraction, and medical education, and investigate whether LLMs possess the transformative power to revolutionize these tasks or whether the distinct complexities of biomedical domain presents unique challenges. Following an extensive literature survey, we find that significant advances have been made in the field of text generation tasks, surpassing the previous state-of-the-art methods. For other applications, the advances have been modest. Overall, LLMs have not yet revolutionized biomedicine, but recent rapid progress indicates that such methods hold great potential to provide valuable means for accelerating discovery and improving health. We also find that the use of LLMs, like ChatGPT, in the fields of biomedicine and health entails various risks and challenges, including fabricated information in its generated responses, as well as legal and privacy concerns associated with sensitive patient data. We believe this survey can provide a comprehensive and timely overview to biomedical researchers and healthcare practitioners on the opportunities and challenges associated with using ChatGPT and other LLMs for transforming biomedicine and health.