Search

Abstract Machine learning (ML) has achieved substantial success in performing healthcare tasks in which the configuration of every part of the ML pipeline relies heavily on technical knowledge. To help professionals with borderline expertise to better use ML techniques, Automated ML (AutoML) has emerged as a prospective solution. However, most models generated by AutoML are black boxes that are challenging to comprehend and deploy in healthcare settings. We conducted a systematic review to examine AutoML with interpretation systems for healthcare. We searched four databases (MEDLINE, EMBASE, Web of Science, and Scopus) complemented with seven prestigious ML conferences (AAAI, ACL, ICLR, ICML, IJCAI, KDD, and NeurIPS) that reported AutoML with interpretation for healthcare before September 1, 2023. We included 118 articles related to AutoML with interpretation in healthcare. First, we illustrated AutoML techniques used in the included publications, including automated data preparation, automated feature engineering, and automated model development, accompanied by a real‐world case study to demonstrate the advantages of AutoML over classic ML. Then, we summarized interpretation methods: feature interaction and importance, data dimensionality reduction, intrinsically interpretable models, and knowledge distillation and rule extraction. Finally, we detailed how AutoML with interpretation has been used for six major data types: image, free text, tabular data, signal, genomic sequences, and multi‐modality. To some extent, AutoML with interpretation provides effortless development and improves users' trust in ML in healthcare settings. In future studies, researchers should explore automated data preparation, seamless integration of automation and interpretation, compatibility with multi‐modality, and utilization of foundation models.

IntroductionDespite popular speculation that nostalgic social media use skyrocketed during the COVID-19 pandemic, this has yet to be formally investigated in the scientific literature.MethodsInterrupted time series analysis (ITSA) using a segmented regression model was performed to examine the changes in the weekly volume of searches for nostalgic songs on Douyin (the Chinese version of TikTok), as a proxy for nostalgic social media use, before and after the lockdown of Wuhan (signaled the start of the pandemic on a national scale in China).ResultsAcross the study period (January 1, 2019–February 28, 2021), an immediate and significant increase in nostalgic social media use was observed when the pandemic initially started (95% CI = [47314.30, 154969.60], p

Abstract Artificial intelligence (AI) has been extensively researched in medicine, but its practical application remains limited. Meanwhile, there are various disparities in existing AI-enabled clinical studies, which pose a challenge to global health equity. In this study, we conducted an in-depth analysis of the geo-economic distribution of 159 AI-enabled clinical studies, as well as the gender disparities among these studies. We aim to reveal these disparities from a global literature perspective, thus highlighting the need for equitable access to medical AI technologies.

BackgroundNonadherence to medication among patients with cardiovascular diseases undermines the desired therapeutic outcomes. eHealth interventions emerge as promising strategies to effectively tackle this issue. ObjectiveThe aim of this study was to conduct a network meta-analysis (NMA) to compare and rank the efficacy of various eHealth interventions in improving medication adherence among patients with cardiovascular diseases (CVDs). MethodsA systematic search strategy was conducted in PubMed, Embase, Web of Science, Cochrane, China National Knowledge Infrastructure Library (CNKI), China Science and Technology Journal Database (Weipu), and WanFang databases to search for randomized controlled trials (RCTs) published from their inception on January 15, 2024. We carried out a frequentist NMA to compare the efficacy of various eHealth interventions. The quality of the literature was assessed using the risk of bias tool from the Cochrane Handbook (version 2.0), and extracted data were analyzed using Stata16.0 (StataCorp LLC) and RevMan5.4 software (Cochrane Collaboration). The certainty of evidence was evaluated using the Grading of Recommendations, Assessment, Development, and Evaluations (GRADE) approach. ResultsA total of 21 RCTs involving 3904 patients were enrolled. The NMA revealed that combined interventions (standardized mean difference [SMD] 0.89, 95% CI 0.22-1.57), telephone support (SMD 0.68, 95% CI 0.02-1.33), telemonitoring interventions (SMD 0.70, 95% CI 0.02-1.39), and mobile phone app interventions (SMD 0.65, 95% CI 0.01-1.30) were statistically superior to usual care. However, SMS compared to usual care showed no statistical difference. Notably, the combined intervention, with a surface under the cumulative ranking curve of 79.3%, appeared to be the most effective option for patients with CVDs. Regarding systolic blood pressure and diastolic blood pressure outcomes, the combined intervention also had the highest probability of being the best intervention. ConclusionsThe research indicates that the combined intervention (SMS text messaging and telephone support) has the greatest likelihood of being the most effective eHealth intervention to improve medication adherence in patients with CVDs, followed by telemonitoring, telephone support, and app interventions. The results of these network meta-analyses can provide crucial evidence-based support for health care providers to enhance patients’ medication adherence. Given the differences in the design and implementation of eHealth interventions, further large-scale, well-designed multicenter trials are needed. Trial RegistrationINPLASY 2023120063; https://inplasy.com/inplasy-2023-12-0063/

This paper studies the impact of digital economy development on carbon emissions, which is a hot topic in the field of environmental economics. This topic is within the scope of the Frontiers in Environmental Science and is particularly in line with the theme of Environmental Economics and Management section in the journal. Firstly, this paper constructs the Digital Economy Development Index and its variable indicators to measure the development of the digital economy. Secondly, this paper uses urban panel data to reveal the impact of digital economy development on carbon emissions, and further adopts IV method to deal with endogeneity issues. Thirdly, this paper performs the heterogeneity analyses depending on the characteristics of the city. The results can provide a basis for formulating differentiated policies. Fourthly, this paper further explores the mechanism by which the digital economy affects carbon peaking. The results show that the digital economy may influence carbon peaking through upgrading of the industrial structure, technical innovation, and energy consumption.

Changes in olfactory function are considered to be early biomarkers of Parkinson’s disease. Olfactory dysfunction is one of the earliest non-motor features of Parkinson’s disease, appearing in about 90% of patients with early-stage Parkinson’s disease, and can often predate the diagnosis by years. Therefore, olfactory dysfunction should be considered a reliable marker of the disease. However, the mechanisms responsible for olfactory dysfunction are currently unknown. In this article, we clearly explain the pathology and medical definition of olfactory function as a biomarker for early-stage Parkinson’s disease. On the basis of the findings of clinical olfactory function tests and animal model experiments as well as neurotransmitter expression levels, we further characterize the relationship between olfactory dysfunction and neurodegenerative diseases as well as the molecular mechanisms underlying olfactory dysfunction in the pathology of early-stage Parkinson’s disease. The findings highlighted in this review suggest that olfactory dysfunction is an important biomarker for preclinical-stage Parkinson’s disease. Therefore, therapeutic drugs targeting non-motor symptoms such as olfactory dysfunction in the early stage of Parkinson’s disease may prevent or delay dopaminergic neurodegeneration and reduce motor symptoms, highlighting the potential of identifying effective targets for treating Parkinson’s disease by inhibiting the deterioration of olfactory dysfunction.

Abstract Artificial intelligence (AI) has demonstrated the ability to extract insights from data, but the fairness of such data-driven insights remains a concern in high-stakes fields. Despite extensive developments, issues of AI fairness in clinical contexts have not been adequately addressed. A fair model is normally expected to perform equally across subgroups defined by sensitive variables (e.g., age, gender/sex, race/ethnicity, socio-economic status, etc.). Various fairness measurements have been developed to detect differences between subgroups as evidence of bias, and bias mitigation methods are designed to reduce the differences detected. This perspective of fairness, however, is misaligned with some key considerations in clinical contexts. The set of sensitive variables used in healthcare applications must be carefully examined for relevance and justified by clear clinical motivations. In addition, clinical AI fairness should closely investigate the ethical implications of fairness measurements (e.g., potential conflicts between group- and individual-level fairness) to select suitable and objective metrics. Generally defining AI fairness as “equality” is not necessarily reasonable in clinical settings, as differences may have clinical justifications and do not indicate biases. Instead, “equity” would be an appropriate objective of clinical AI fairness. Moreover, clinical feedback is essential to developing fair and well-performing AI models, and efforts should be made to actively involve clinicians in the process. The adaptation of AI fairness towards healthcare is not self-evident due to misalignments between technical developments and clinical considerations. Multidisciplinary collaboration between AI researchers, clinicians, and ethicists is necessary to bridge the gap and translate AI fairness into real-life benefits.

Although the triglyceride form is the natural form of fish oil found in fish, the ethyl ester form of fish oil, which is used during processing to save costs, is also present on the market. In this study, fatty acids and lipids were determined using gas chromatography–mass spectrometry (GC–MS) and liquid chromatography–linear ion trap mass spectrometry (LC–LIT/MS), respectively, according to developed methods. The identification of fatty acids was based on the mass spectral characteristics and equivalent chain lengths. However, the fatty acid contents of both forms of fish oils are quite similar. The application of the LC–LIT/MS method for the structural characterization of triacylglycerols (TAGs) and the mechanism of LIT/MS fragmentation are also discussed. Neutral losses of CH2=CH2 (m/z 28) and CH3CH2OH (m/z 46), which are LIT/MS characteristics of ethyl ester from fish oil, were found for the first time. The triglyceride form of fish oils was easily and accurately identified using fingerprint chromatography. In conclusion, lipid analysis combined with LC–LIT/MS showed an improved capability to distinguish between types of fish oil.

Purpose – With the increasing requirements for intelligence in the field of aviation manufacturing, manual assembly can hardly adapt to the trend of future production. The purpose of this study is to realize the semi-automatic assembly of the movable airfoil by proposing a human-robot collaborative assembly strategy based on adaptive admittance control. Design/methodology/approach – A logical judgment system for operating intentions is introduced in terms of different situations of the movements; hence, a human cognition-based adaptive admittance control method is developed to curb the damage of inertia; then virtual limit walls are raised on the periphery of the control model to ensure safety; finally, simulated and experimental comparisons with other admittance control methods are conducted to validate the proposed method. Findings – The proposed method can save at least 28.8% of the time in the stopping phase which effectively compensates for inertia during the assembly process and has high robustness concerning data disturbances. Originality/value – Due to the human-robot collaboration to achieve compliant assembly of movable airfoils can preserve human subjectivity while overcoming the physical limits of humans, which is of great significance to the investigation of intelligent aircraft assembly, the proposed method that reflects the user's naturalness and intuitiveness can not only enhance the stability and the flexibility of the manipulation, but also contribute to applications of industrial robots in the field of human-robot collaboration.

Abstract Designing and synthesizing advanced electrocatalysts with superior intrinsic activity toward hydrogen evolution reaction (HER) in alkaline media is critical for the hydrogen economy. Herein, a novel Ir@Rhene heterojunction electrocatalyst is synthesized via epitaxially confining ultrasmall and low‐coordinate Ir nanoclusters on the ultrathin Rh metallene accompanying the formation of Ir/IrO2 Janus nanoparticles. The as‐prepared heterojunctions display outstanding alkaline HER activity, with an overpotential of only 17 mV at 10 mA cm−2 and an ultralow Tafel slope of 14.7 mV dec−1. Both structural characterizations and theoretical calculations demonstrate that the Ir@Rhene heterointerfaces induce charge density redistribution, resulting in the increment of the electron density around the O atoms in the IrO2 site and thus delivering much lower water dissociation energy. In addition, the dual‐site synergetic effects between IrO2 and Ir/Rh interface trigger and improve the interfacial hydrogen spillover, thereby subtly avoiding the steric blocking of the active site and eventually accelerating the alkaline HER kinetics.

Summary: The AutoScore framework can automatically generate data-driven clinical scores in various clinical applications. Here, we present a protocol for developing clinical scoring systems for binary, survival, and ordinal outcomes using the open-source AutoScore package. We describe steps for package installation, detailed data processing and checking, and variable ranking. We then explain how to iterate through steps for variable selection, score generation, fine-tuning, and evaluation to generate understandable and explainable scoring systems using data-driven evidence and clinical knowledge.For complete details on the use and execution of this protocol, please refer to Xie et al. (2020),1 Xie et al. (2022)2, Saffari et al. (2022)3 and the online tutorial https://nliulab.github.io/AutoScore/. : Publisher’s note: Undertaking any experimental protocol requires adherence to local institutional guidelines for laboratory safety and ethics.

In this paper, we explore the problem of direction-of-arrival (DOA) estimation for a non-uniform linear array (NULA) under strong noise. The compressed sensing (CS)-based methods are widely used in NULA DOA estimations. However, these methods commonly rely on the tuning of parameters, which are hard to fine-tune. Additionally, these methods lack robustness under strong noise. To address these issues, this paper proposes a novel DOA estimation approach using a deep neural network (DNN) for a NULA in a low SNR. The proposed network is designed based on the denoising convolutional neural network (DnCNN) and the alternating direction method of multipliers (ADMM), which is dubbed as LDnADMM-Net. First, we construct an unfolded DNN architecture that mimics the behavior of the iterative processing of an ADMM. In this way, the parameters of an ADMM can be transformed into the network weights, and thus we can adaptively optimize these parameters through network training. Then, we employ the DnCNN to develop a denoising module (DnM) and integrate it into the unfolded DNN. Using this DnM, we can enhance the anti-noise ability of the proposed network and obtain a robust DOA estimation in a low SNR. The simulation and experimental results show that the proposed LDnADMM-Net can obtain high-accuracy and super-resolution DOA estimations for a NULA with strong robustness in a low signal-to-noise ratio (SNR).

To ameliorate the precision of clothing image classification, we proposed a clothing image classification method via the DenseNet201 network based on transfer learning and the optimized regularized random vector functional link (RVFL). First, the formula extracts weight’s parameters about DenseNet201 that is pre-trained on the ImageNet dataset for transfer learning, thereby obtaining an incipient network，after that trim this model parameters. The modified network is utilized to pick up the clothing image features output by the DenseNet201’s global average pooling layer. Second, regularization coefficient is introduced to control RVFL’s model complexity and solve the problem of over-fitting. Then, the generated solution vector of aquila optimizer (AO) is produced by marine predators algorithm (MPA). The input weights, biases of hidden layer and renormalization modulus of regularized RVFL are optimized using the improved AO algorithm. Finally, we use the optimized RVFL to assort abstracted fashion graphics traits. We use Accuracy, Macro-F1, Macro-R and Macro-P to assess the algorithm’s ability and compare this algorithm with ResNet50 network, ResNet101 network, DenseNet201 network, InceptionV3 network and different classifiers, which use DenseNet201 as the feature extractor to get the input. From the experimental results, this algorithm proposed has excellent classification power and generalization ability.

BackgroundTraditional Chinese medicines (TCMs), such as Tripterygium wilfordii Hook F (TwHF), Glycyrrhiza uralensis, Caulis sinomenii and others have anti-inflammatory effects. They are widely used in China to treat rheumatoid arthritis (RA), but proof of their use as an evidence-based medicine is little. The aim of this network meta-analysis (NMA) was to evaluate the efficacy and safety of TCMs.MethodsBy searching online databases and using a manual retrieval method, randomized controlled trials (RCTs) that met specific selection criteria were included in the meta-analysis. The search included papers that were published between the establishment of the databases and November 10, 2022. Analyses were performed using Stata software (version 14) and Review Manager (version 5.3).Results61 papers with 6316 subjects were included in the current NMA. For ACR20, MTX plus SIN therapy (94.30%) may be a significant choice. For ACR50 and ACR70, MTX plus IGU therapy (95.10%, 75.90% respectively) performed better than other therapies. IGU plus SIN therapy (94.80%) may be the most promising way to reduce DAS-28, followed by MTX plus IGU therapy (92.80%) and TwHF plus IGU therapy (83.80%). In the analysis of the incidence of adverse events, MTX plus XF therapy (92.50%) had the least potential, while LEF therapy (22.10%) may cause more adverse events. At the same time, TwHF therapy, KX therapy, XF therapy and ZQFTN therapy were not inferior to MTX therapy.ConclusionsTCMs with anti-inflammatory effect were not inferior to MTX therapy in the treatment of RA patients. Combining with TCMs can improve the clinic efficacy and reduce the possibility of adverse events of DMARDs, which may be a promising regimen.Systematic review registrationhttps://www.crd.york.ac.uk/PROSPERO/, identifier CRD42022313569.

Diabetic wounds are a common disease that plague many doctors in the clinic. Diabetic wounds are caused by increased blood sugar levels, which leads to microvascular disease and skin dystrophy. Additionally, their occurrence can be accompanied by inflammation and a variety of bacterial infections. Diabetic wounds are complex; therefore, achieving good results after a single treatment modality is often difficult. In this study, we used materials science engineering methods to integrate several factors required for diabetic wound treatment, which resulted in the development of a bilayer hydrogel. First, the upper layer of the hydrogel inhibits bacterial growth and promotes wound closing, which solves the problems of wound infection and protection. Additionally, the lower layer of the hydrogel has a porous structure, which promotes the adhesion and growth of wound healing-related cells. Moreover, the lower layer of the hydrogel can release cytokines that promote vascular endothelial cell and fibroblast migration and stimulate skin tissue regeneration. We evaluated the ability of this double-layer hydrogel to heal diabetic wounds in vivo and in vitro and found that it offered protection while promoting healing. We also determined that a combination of multiple factors can provide better reparative effects to treat chronic diabetic wounds.

Summary: Background: Return of spontaneous circulation (ROSC) before arrival at the emergency department is an early indicator of successful resuscitation in out-of-hospital cardiac arrest (OHCA). Several ROSC prediction scores have been developed with European cohorts, with unclear applicability in Asian settings. We aimed to develop an interpretable prehospital ROSC (P-ROSC) score for ROSC prediction based on patients with OHCA in Asia. Methods: This retrospective study examined patients who suffered from OHCA between Jan 1, 2009 and Jun 17, 2018 using data recorded in the Pan-Asian Resuscitation Outcomes Study (PAROS) registry. AutoScore, an interpretable machine learning framework, was used to develop P-ROSC. On the same cohort, the P-ROSC was compared with two clinical scores, the RACA and the UB-ROSC. The predictive power was evaluated using the area under the curve (AUC) in the receiver operating characteristic analysis. Findings: 170,678 cases were included, of which 14,104 (8.26%) attained prehospital ROSC. The P-ROSC score identified a new variable, prehospital drug administration, which was not included in the RACA score or the UB-ROSC score. Using only five variables, the P-ROSC score achieved an AUC of 0.806 (95% confidence interval [CI] 0.799–0.814), outperforming both RACA and UB-ROSC with AUCs of 0.773 (95% CI 0.765–0.782) and 0.728 (95% CI 0.718–0.738), respectively. Interpretation: The P-ROSC score is a practical and easily interpreted tool for predicting the probability of prehospital ROSC. Funding: This research received funding from SingHealth Duke-NUS ACP Programme Funding (15/FY2020/P2/06-A79).

Over the past two decades, a number of advances in topic modeling have produced sophisticated models that are capable of generating topic hierarchies. In particular, hierarchical Latent Dirichlet Allocation (hLDA) builds a topic tree based on the nested Chinese Restaurant Process (nCRP) or other sampling processes to generate a topic hierarchy that allows arbitrarily large branch structures and adaptive dataset growth. In addition, hierarchical topic models based on the latent tree model, such as Hierarchical Latent Tree Analysis (HLTA), have been developed over the last five years. However, these models do not work well in cases with millions of documents and hundreds of thousands of terms. In addition, the topic trees generated by these models are always poorly interpretable, and the relationships among topics in different levels are relatively simple. The biomedical literature, including Medline abstracts, has large-scale documents in two major categories: biological laboratory research and medical clinical research. We propose a top-down binary hierarchical topic model (biHTM) for biomedical literature by iteratively applying a flat topic model and adaptively processing subtrees of the hierarchy. The biHTM topic hierarchy of complete Medline abstracts with more than 14 topic node levels shows good bimodality and interpretability. Compared to hLDA and HLTA, biHTM shows promising results in experiments assessed in terms of runtime and quality.

Coronary heart disease (CHD) is one of the main causes of death worldwide. In the past few decades, several in-depth research on the pathological mechanisms and effective treatment methods for CHD have been conducted. At present, the intervention of a variety of therapeutic drugs and treatment technologies have greatly reduced the burden on global public health. However, severe arrhythmia and myocardial fibrosis accompanying CHD in the later stages need to be addressed urgently. Mitochondria are important structural components for energy production and the main sites for aerobic respiration in cells. Mitochondria are involved in arrhythmia, myocardial fibrosis, and acute CHD and play a crucial role in regulating myocardial ischemia/hypoxia. Mitochondrial dysfunction or mitophagy disorders (including receptor-dependent mitophagy and receptor-independent mitophagy) play an important role in the pathogenesis of CHD, especially mitophagy. Mitophagy acts as a “mediator” in the inflammatory damage of cardiomyocytes or vascular endothelial cells and can clear mitochondria or organelles damaged by inflammation under normal conditions. We reviewed experimental advances providing evidence that mitochondrial homeostasis or mitochondrial quality control are important in the pathological mechanism of CHD. Further, we reviewed and summarized relevant regulatory drugs that target mitochondrial function and quality control.

In order to understand the effect of digital marketing on brand visibility of wineries in the age of internet 2.0, the researchers conducted a qualitative research to five Napa Valley wineries. Based on the primary and secondary data collected through the communication between the researchers and the wineries via email, home website, and different platforms of social media, a behavioral model proposed by Capitello (2013) was introduced to examine the effectiveness of digital marketing towards brand visibility. Some of the parameters and dimensions in the model were adjusted to fit the appropriateness of the study. As a result, communication style impacts significantly on consumer’s first impression to the company, and the wineries need to allocate the resources wisely. The researchers also proposed a further study about the relationship between the brand recognition and online marketing efforts under a sufficient period and similar brand positioning so as to have a fuller picture about the topic.

BackgroundDeath and disability from coronary heart disease (CHD) can be largely reduced by improving risk factor management. However, adhering to evidence-based recommendations is challenging and requires interventions at the level of the patient, provider, and health system. ObjectiveThe aim of this study was to develop an Intelligent Individualized Cardiovascular App for Risk Elimination (iCARE) to facilitate adherence to health behaviors and preventive medications, and to test the usability of iCARE. MethodsWe developed iCARE based on a user-centered design approach, which included 4 phases: (1) function design, (2) iterative design, (3) expert inspections and walkthroughs of the prototypes, and (4) usability testing with end users. The usability testing of iCARE included 2 stages: stage I, which included a task analysis and a usability evaluation (January to March 2019) of the iCARE patient app using the modified Health Information Technology Usability Survey (Health-ITUES); and stage II (June 2020), which used the Health-ITUES among end users who used the app for 6 months. The end users were individuals with a confirmed diagnosis of CHD from 2 university-affiliated hospitals in Beijing, China. ResultsiCARE consists of a patient app, a care provider app, and a cloud platform. It has a set of algorithms that trigger tailored feedback and can send individualized interventions based on data from initial assessment and health monitoring via manual entry or wearable devices. For stage I usability testing, 88 hospitalized patients (72% [63/88] male; mean age 60 [SD 9.9] years) with CHD were included in the study. The mean score of the usability testing was 90.1 (interquartile range 83.3-99.0). Among enrolled participants, 90% (79/88) were satisfied with iCARE; 94% (83/88) and 82% (72/88) reported that iCARE was useful and easy to use, respectively. For stage II usability testing, 61 individuals with CHD (85% [52/61] male; mean age 53 [SD 8.2] years) who were from an intervention arm and used iCARE for at least six months were included. The mean total score on usability testing based on the questionnaire was 89.0 (interquartile distance: 77.0-99.5). Among enrolled participants, 89% (54/61) were satisfied with the use of iCARE, 93% (57/61) perceived it as useful, and 70% (43/61) as easy to use. ConclusionsThis study developed an intelligent, individualized, evidence-based, and theory-driven app (iCARE) to improve patients’ adherence to health behaviors and medication management. iCARE was identified to be highly acceptable, useful, and easy to use among individuals with a diagnosis of CHD. Trial RegistrationChinese Clinical Trial Registry ChiCTR-INR-16010242; https://tinyurl.com/2p8bkrew

Avatars or agents are digitized self-representations of a player in mediated environments. While using agents to navigate through mediated environments, players form bonds with their self-agents or characters, a process referred to as identification. Identification can involve automatic, but temporary, self-concept “shifts in implicit self-perceptions” (Klimmt et al., 2010, p. 323) of the media user by adopting or emphasizing the action choices on behalf of the social expectation of the avatar in the mediated environment. In the current study, we test the possibility that users' identification with video game avatars–a bond built between avatars and players- would account for subsequent behavior changes. We did so by using 3-month longitudinal data involving a narratively-based serious game: Socially Optimized Learning in Virtual Environments (SOLVE), a 3D-interactive game designed to reduce risky sexual behaviors among young men who have sex with men (n = 444). Results show that video game identification predicts both the reduction of risky sexual behaviors over time, and reduction in the number of non-primary partners with whom risky sex occurs. And when players identify with the game character, they tend to make healthier choices, which significantly mediates the link between video game identification and reduction of risky behaviors.

Background: The performance of GPT-4 in nursing examinations within the Chinese context has not yet been thoroughly evaluated. Objective: To assess the performance of GPT-4 on multiple-choice and open-ended questions derived from nursing examinations in the Chinese context. Methods: The data sets of the Chinese National Nursing Licensure Examination spanning 2021 to 2023 were used to evaluate the accuracy of GPT-4 in multiple-choice questions. The performance of GPT-4 on open-ended questions was examined using 18 case-based questions. Results: For multiple-choice questions, GPT-4 achieved an accuracy of 71.0% (511/720). For open-ended questions, the responses were evaluated for cosine similarity, logical consistency, and information quality, all of which were found to be at a moderate level. Conclusion: GPT-4 performed well at addressing queries on basic knowledge. However, it has notable limitations in answering open-ended questions. Nursing educators should weigh the benefits and challenges of GPT-4 for integration into nursing education. [ABSTRACT FROM AUTHOR]