Evaluating large language models in analysing classroom dialogue.

This study explores the use of Large Language Models (LLMs), specifically GPT-4, in analysing classroom dialogue-a key task for teaching diagnosis and quality improvement. Traditional qualitative methods are both knowledge- and labour-intensive. This research investigates the potential of LLMs to streamline and enhance this process. Using datasets from middle school mathematics and Chinese classes, classroom dialogues were manually coded by experts and then analysed with a customised GPT-4 model. The study compares manual annotations with GPT-4 outputs to evaluate efficacy. Metrics include time efficiency, inter-coder agreement, and reliability between human coders and GPT-4. Results show significant time savings and high coding consistency between the model and human coders, with minor discrepancies. These findings highlight the strong potential of LLMs in teaching evaluation and facilitation.

Unloading and successful treatment with bioresorbable stents during percutaneous coronary intervention: A case report.

BACKGROUND: With the development of percutaneous coronary intervention (PCI), the number of interventional procedures without implantation, such as bioresorbable stents (BRS) and drug-coated balloons, has increased annually. Metal drug-eluting stent unloading is one of the most common clinical complications. Comparatively, BRS detachment is more concealed and harmful, but has yet to be reported in clinical research. In this study, we report a case of BRS unloading and successful rescue. CASE SUMMARY: This is a case of a 59-year-old male with the following medical history: "Type 2 diabetes mellitus" for 2 years, maintained with metformin extended-release tablets, 1 g PO BID; "hypertension" for 20 years, with long-term use of metoprolol sustained-release tablets, 47.5 mg PO QD; "hyperlipidemia" for 20 years, without regular medication. He was admitted to the emergency department of our hospital due to intermittent chest pain lasting 18 hours, on February 20, 2022 at 15: 35. Electrocardiogram results showed sinus rhythm, ST-segment elevation in leads I and avL, and poor R-wave progression in leads V1-3. High-sensitivity troponin I level was 4.59 ng/mL, indicating an acute high lateral wall myocardial infarction. The patient's family requested treatment with BRS, without implantation. During PCI, the BRS became unloaded but was successfully rescued. The patient was followed up for 2 years; he had no episodes of angina pectoris and was in generally good condition. CONCLUSION: We describe a case of a 59-year-old male experienced BRS unloading and successful rescue. By analyzing images, the causes of BRS unloading and the treatment plan are discussed to provide insights for BRS release operations. We discuss preventive measures for BRS unloading.

Peripheral perfusion index of pulse oximetry in adult patients: a narrative review.

The peripheral perfusion index (PI) is derived from pulse oximetry and is defined as the ratio of the pulse wave of the pulsatile portion (arteries) to the non-pulsatile portion (venous and other tissues). A growing number of clinical studies have supported the use of PI in various clinical scenarios, such as guiding hemodynamic management and serving as an indicator of outcome and organ function. In this review, we will introduce and discuss this traditional but neglected indicator of the peripheral microcirculatory perfusion. Further clinical trials are required to clarify the normal and critical values of PI for different monitoring devices in various clinical conditions, to establish different standards of PI-guided strategies, and to determine the effect of PI-guided therapy on outcome.

Efficacy and safety of early prone position in postoperative cardiac surgery adults with acute respiratory distress syndrome: a single-center retrospective cohort study.

Background: Acute respiratory distress syndrome (ARDS) is a leading cause of postoperative respiratory failure after cardiac surgery, and the mortality rate is extremely high. Although prone positioning (PP) may be safe and effective for ARDS, it is still not widely adopted in cardiac surgery patients. We aimed to assess the efficacy and safety of early PP in ARDS after cardiac surgery. Methods: This is a single-center retrospective cohort study. We included adult intensive care unit (ICU) patients who developed ARDS with arterial pressure of oxygen to fraction of oxygen ratio (P/F) ≤200 mmHg within 72 hours after cardiac surgery between 1 January 2019 and 1 August 2023. The outcomes were P/F after 1 session of PP, duration of mechanical ventilation (MV) and ICU stay, and adverse events. Results: In total, 79 patients who underwent PP and 87 patients who underwent supine position (SP) were included. The mean time to perform PP after ICU admission was 38.0 hours. The P/F improved significantly after 1 session of PP treatment [160.0 (127.8-184.3) vs. 275.0 (220.0-325.0) mmHg, P<0.001], the duration of MV and ICU stay in the PP group were significantly shorter than those in the SP group [84.0 (64.0-122.0) vs. 120.0 (97.0-182.0) h, P<0.001; 6.0 (5.0-8.0) vs. 8.0 (6.0-12.0) days, P<0.001, respectively]. No adverse events were observed during the PP even in patients with intra-aortic balloon pump (IABP). Conclusions: Early PP treatment is effective and safe for patients with moderate to severe ARDS after cardiac surgery and it is even safe in a subgroup placed with IABP.

"Application of CT radiomics in brain metastasis of lung cancer: A systematic review and meta-analysis".

PURPOSE: This study aimed to systematically assess the quality and performance of computed tomography (CT) radiomics studies in predicting brain metastasis (BM) among patients with lung cancer. METHODS: The PubMed, Embase and Web of Science were searched for studies predicting BM in patients with lung cancer using CT-based radiomics features. Information regarding patients, imaging, and radiomics analysis was extracted from eligible studies. We assessed the quality of included studies using the Radiomics Quality Scoring (RQS) tool and the Quality Assessment of Diagnostic Accuracy Studies (QUADAS-2). A meta-analysis of studies regarding the prediction of BM in patients with lung cancer was performed. RESULTS: Thirteen studies were identified, with sample sizes ranging from 75 to 602. The mean RQS of the studies was 12 (range 9-16), and the corresponding percentage of the score was 33.55 % (range 25.00-44.44 %). Four studies (30.8 %) were considered as low risk of bias, while the remaining nine studies (69.2 %) were considered to have unclear risks. The meta-analysis included twelve studies. The pooled sensitivity, specificity and Area Under the Curve (AUC) value with 95 % confidence intervals were 0.75 [0.69, 0.80], 0.76 [0.68, 0.82], and 0.81 [0.77-0.84], respectively. CONCLUSION: CT radiomics-based models show promising results as a non-invasive method to predict BM in lung cancer patients. However, multicenter and prospective studies are warranted to enhance the stability and acceptance of radiomics.

Full connected layer model with self-attention to hourly predict heparin dosage for perioperative cardiac surgery patients.

BACKGROUND: Anticoagulant therapy for patients who underwent cardiac surgery is a challenge. Both inadequate and excessive anticoagulation can cause fatal complications. Previous studies failed to provide real-time guidance for heparin pump speed adjustment. This study intended to provide a quantitative prediction model to optimize heparin dosage for cardiac surgery patients based on machine learning method. METHODS: Patients who underwent cardiac surgery and admitted to intensive care unit in Peking Union Medical College Hospital (PUMCH) from January 2013 to December 2023 were retrospectively analyzed. In order to reach target activated partial thromboplastin time (aPTT), linear regression, SVM, XGBoost, LSTM, GRU, FC (Full Connected Layer) and FC + self-attention models were used to make hourly adjustment recommendation for administrations of heparin pump speed. Mean absolute square, and absolute percentage errors were used to evaluate the reliability of the models. SHAP method and feature cumulative effect were used to interpret the features of the FC + self-attention model. Safety and economic evaluation based on clinical compliance of this real-world data-oriented model was further analyzed. RESULTS: A total of 1080 patients including 112,554 heparin pump administrations were included in this study. Among seven candidate models, FC + self-attention model yielded the lowest mean absolute error of 0.9388 and 1.1325 in test and validation cohort. Gap to target aPTT, thrombin time, history of coronary heart disease, previous duration of arterial fibrillation and prothrombin activity were identified as important features affecting heparin adjustment. High compliance to FC + self-attention model may increase percentage of normal therapeutic time and decrease supratherapeutic therapeutic time and reducing blood draw until two consecutive normal therapeutic stabilization of aPTT. CONCLUSIONS: This FC + self-attention model is potentially applicable for giving recommendation for healthcare providers to optimize heparin dosage for cardiac surgery patients.

Complement system activation: bridging physiology, pathophysiology, and therapy.

The complement system is a set of over 50 proteins that constitutes an essential part of the innate immune system. Complement system activation involves an organized proteolytic cascade. Overactivation of complement system activation is the main pathogenic mechanism of several diseases and contributes to the manifestations of many other conditions. This review describes the normal complement system and the role for complement dysregulation in critical illnesses, notably sepsis and acute respiratory distress syndrome. Complement activation is involved in the immune system response to pathogens but, when excessive, can contribute to tissue damage, runaway inflammation, and capillary leakage syndrome. Complement overactivation may play a key role in severe forms of coronavirus disease 2019 (COVID-19). Two diseases whose manifestations are mainly caused by complement overactivation, namely, atypical hemolytic and uremic syndrome (aHUS) and myasthenia gravis, are discussed. A diagnostic algorithm for aHUS is provided. Early complement-inhibiting therapy has been proven effective. When renal transplantation is required, complement-inhibiting drugs can be used prophylactically to prevent aHUS recurrence. Similarly, acetylcholine-receptor autoantibody-positive generalized myasthenia gravis involves complement system overactivation and responds to complement inhibition. The two main complement inhibitors used in to date routine are eculizumab and ravulizumab. The main adverse event is Neisseria infection, which is rare and preventable, but can be fatal. The complement system is crucial to health but, when overactivated, can cause or contribute to disease. Effective complement inhibitors are now available, although additional data are required to determine optimal regimens. Further research is also needed to better understand the complement system, develop advanced diagnostic tools, and identify markers that allow the personalization of treatment strategies.

Response of anammox to organics with different degradation characteristics and exposure time: Performance, sludge characteristics and bacterial community.

The effects of typical organic compounds including easily degradable organic matters sodium acetate, yeast and methanol, and refractory organic matter (ROM) humic acid on anaerobic ammonium oxidation (anammox) systems in short-term and medium-term exposure time were studied. During short-term experiments, nitrogen removal activity (NRA) was inhibited at sodium acetate level of 150 mg L-1 total organic carbon (TOC) and methanol level of 30-150 mg L-1 TOC, but humic acid and yeast (≤150 mg L-1 TOC) enhanced nitrogen removal in anammox systems. The greatest NRA of 30.10 mg TN g-1 VSS h-1 was recorded at yeast level of 90 mg L-1 TOC. In medium-term experiments, organics significantly inhibited the nitrogen removal ability. As a ROM, humic acid enhanced sludge aggregation and biological diversity, but decreased the bioactivity and extracellular polymeric substances levels. Due to the endogenous denitrification, the relative abundance of anammox bacteria (AnAOB) was decreased. Candidatus Kuenenia is still dominant in sludge with methanol and humid acid, but AnAOB are not dominant due to the addition of sodium acetate and yeast. This research would be beneficial for the full-scale application of the anammox process in treating real wastewater with organics and ammonia.

High-Entropy Fluorite Oxide-Modified CaO-Based Sorbent Pellets for Enhanced High-Temperature CO2 Capture.

The calcium looping technology employing CaO-based sorbents is pivotal for capturing CO2 from flue gas. However, the intrinsic low thermodynamic stability of CaO-based sorbents and the requisite molding step induce severe sintering issues, diminishing their cyclic stability. Herein, a high-entropy fluorite oxide (HEFO) inert stabilizer premised on entropy stabilization and synergistic effect strategies is introduced. HEFO-modified, CaO-based sorbent pellets are synthesized via a rapid cigarette butt-assisted combustion process (15 min) combined with the graphite molding method. Post-multiple cycles, their CO2 capture capacity reaches 0.373 g g-1, which is 2.6-fold superior to that of pure CaO, demonstrating markedly enhanced anti-sintering properties. First, the subtle morphological and crystallographic modifications suggest that the inherent entropy stability of HEFO imparts robust thermal resistance. Concurrently, the disordered structure of single-phase HEFO exhibits a high affinity for CaO, resulting in an interface binding energy of -1.83 eV, in sharp contrast to the -0.112 eV of pure CaO, thereby restricting CaO migration. Additionally, the multi-element synergistic effect of HEFO reduces the energy barrier by 0.15 eV, leading to a 40% and 140% increase in carbonation and calcination rates, respectively. This work presents highly efficient and rapidly synthesized CaO-based sorbent pellets, showcasing promising potential for industrial application.

Cu Nanowires Encapsulated by ZIF67 for Efficient Ammonia Electrosynthesis from Nitrate.

Electrochemical NO3- reduction reaction (NO3RR) represents a green and sustainable way to produce valuable NH3 for both NH3 production and nitrate contaminant removal, and developing efficient, durable, highly selective catalyst is the key. Herein, we report a facile method to fabricate a catalyst composed of ultrafine Cu nanowires (Cu NWs) encapsulated by ZIF67, namely, CuNW@ZIF67, for efficient NH3 electrosynthesis from nitrate. The CuNW@ZIF67 catalyst exhibited excellent catalytic performance toward NO3RR in alkaline electrolyte, manifested by a large NH3 Faradaic efficiency of 93.7% at -0.5 V versus reversible hydrogen electrode (RHE), a high energy efficiency over 30% at -0.7 V, and robust long-term stability. Such intriguing catalytic properties are mainly ascribed to its structural merits and the strong electronic interaction between Cu NWs and ZIF67. DFT calculations revealed that, the Cu site can easily convert NO3- into NO2-, while the Co site plays a critical role in catalyzing the NO2--to-NH3 process. The study can shed light on rational design of efficient, durable, and highly selective catalysts for NO3RR and beyond.

The level of partial pressure of carbon dioxide affects organ perfusion in respiratory failure patients undergoing pressure support ventilation with venovenous extracorporeal membrane oxygenation: a prospective study.

BACKGROUND: We evaluated the influence of different partial carbon dioxide pressure (PaCO2) levels on organ perfusion in patients with respiratory failure receiving pressure-support ventilation with veno-venous extracorporeal membrane oxygenation (V-V ECMO). METHODS: In this twelve patients prospective study, ECMO gas-flow was decreased from baseline (PaCO2 < 40 mmHg) until PaCO2 increased by 5-10 mmHg (High-CO2 phase). Resistance indices of gut, spleen, and snuffbox artery, the peripheral perfusion index (PPI), and heart rate variability were measured at baseline and High-CO2 phase. RESULTS: When PaCO2 increased from 36 (36-37) mmHg at baseline to 42 (41-43) mmHg in the High-CO2 phase (p < 0.001), PPI decreased significantly (p = 0.026). The snuffbox artery (p = 0.022), superior mesenteric artery (p = 0.042), and spleen (p = 0.012) resistance indices increased significantly. The root mean square of successive differences (RMSSD) decreased from 19.5(18.1-22.7) to 15.9(14.4-18.6) ms (p = 0.034), and the ratio of low-frequency to high-frequency components(LF/HF) increased from 0.47 ± 0.23 to 0.70 ± 0.38 (p = 0.013). CONCLUSIONS: High PaCO2 might cause decreased peripheral tissue and visceral organ perfusion through autonomic nervous system in patients with respiratory failure undergoing PSV with V-V ECMO.

Biodegradable Poly (L-Lactic acid) Fibrous Membrane with Ribbon-Structured Fibers and Ultrafine Nanofibers Enhances Air Filtration Performance.

Here, the poly (l-lactic acid) (PLLA) membrane with multi-structured networks (MSN) is successfully prepared by electrospinning technology for the first time. It is composed of micron-sized ribbon-structured fibers and ultrafine nanofibers with a diameter of tens of nanometers, and they are connected to form the new network structure. Thanks to the special fiber morphology and structure, the interception and electrostatic adsorption ability for against atmospheric particulate matter (PM) are significantly enhanced, and the resistance to airflow is reduced due to the "slip effect" caused by ultrafine nanofibers. The PLLA MSN membrane shows excellent filtration performance with ultra-high filtration efficiency (>99.9% for PM2.5 and >99.5% for PM0.3) and ultra-low pressure drop (≈20 Pa). It has demonstrated filtration performance that even exceeds current non-biodegradable polymer materials, laying the foundation for future applications of biodegradable PLLA in the field of air filtration. In addition, this new structure also provides a new idea for optimizing the performance of other polymer materials.

Smart Bandage Based on a ZIF-8 Triboelectric Nanogenerator for In Situ Real-Time Monitoring of Drug Concentration.

For chronic wounds, frequent replacement of bandages not only increases the likelihood of secondary damage and the risk of cross infection but also wastes medication. Therefore, in situ real-time monitoring of the concentrations of residual drugs in bandages is crucial. Here, we propose a novel strategy that combines a triboelectric nanogenerator (TENG) with medical bandages to develop a smart bandage based on zeolite imidazolate framework TENG. During the process of wound healing, the electrical output of TENG changes with the continuous release of drugs. Based on the correlation between the electrical signal of TENG and drug concentration, the concentration of residual drugs in the bandage can be monitored in real-time in situ, guiding medical staff to replace the bandage at the most appropriate time. The smart bandage based on TENG provides a new strategy for in situ real-time monitoring of drug concentration and also provides an ideal and feasible solution for the field of biomedical drug sensing.

Application of a cloud platform that identifies patient-ventilator asynchrony and enables continuous monitoring of mechanical ventilation in intensive care unit.

Background: Patient-ventilator asynchrony (PVA) frequently occurs in mechanically ventilated patients within the ICU and has the potential for harm. Depending solely on the health care team cannot accurately and promptly identify PVA. To address this issue, our team has developed a cloud-based platform for monitoring mechanical ventilation (MV), comprising the PVA-RemoteMonitor system and the 24-h MV analysis report. We conducted a survey to evaluate physicians' satisfaction and acceptance of the platform in 14 ICUs. Methods: Data from medical records, clinical information systems, and ventilators were uploaded to the cloud platform and underwent data processing. The data were analyzed to monitor PVA and displayed in the front-end. The 24-h analysis report for MV was generated for clinical reference. Critical care physicians in 14 hospitals' ICUs that involved in the platform participated in a questionnaire survey, among whom 10 physicians were interviewed to investigate physicians' acceptance and opinions of this system. Results: The PVA-RemoteMonitor system exhibited a high level of specificity in detecting flow insufficiency, premature cycle, delayed cycle, reverse trigger, auto trigger, and overshoot, with sensitivities of 90.31 %, 98.76 %, 99.75 %, 99.97 %, 100 %, and 99.69 %, respectively. The 24-h analysis report supplied essential data about PVA and respiratory mechanics. 86.2 % (75/87) of physicians supported the application of this platform. Conclusions: The PVA-RemoteMonitor system accurately identified PVA, and the MV analysis report provided guidance in controlling PVA. Our platform can effectively assist ICU physicians in the management of ventilated patients.

Co-administration of 1,25-dihydroxyvitamin D3 and infliximab improves colitis in mice by modulating Treg differentiation.

Objectives: The combination of TNF-α inhibitors and vitamin D in colitis remains to be elucidated. In the present study, we revealed the benefit of infliximab (IFX) and vitamin D in a mouse model of Ulcerative colitis (UC). Materials and Methods: A dextran sulfate sodium-induced colitis model was used. The therapeutic effect of the combination was evaluated by symptom and histopathology analysis. The synergistic mechanism was explored by detecting the regulatory effect of the combined therapy on Regulatory T cell (Treg) differentiation. Results: IFX and 1,25-dihydroxyvitamin D3 (VitD3) synergistically prevented the development of colitis by improving clinical signs, pathological and hematological manifestation, and inhibiting intestinal inflammation (decreasing TNF-α, IL-1ß, and IL-6). Co-administration of IFX (2.5 mg/kg) with VitD3 or IFX (5.0 mg/kg) with VitD3 was more effective than administration of IFX (2.5 mg/kg, 5.0 mg/kg). There was no difference in therapeutic effect between IFX (5.0 mg/kg) and VitD3+ IFX (2.5 mg/kg) groups or between the VitD3+IFX (5.0 mg/kg) and VitD3+ Azathioprine (AZA) groups. VitD3 or combination therapy showed more powerful regulation of splenetic Treg differentiation and IL-10 production than IFX alone. Moreover, VitD3 alone or in combination induced higher levels of Foxp3 and IL-10 than IFX in colon tissue. In ulcerative colitis patients, serum VitD3 levels positively correlated with Treg levels. Conclusion: VitD3 and IFX synergistically inhibit colitis based on their powerful regulation of Treg differentiation. VitD3 combined with IFX is an alternative therapy for patients who are intolerant to standard doses of IFX or combination of IFX and AZA.

Circular RNA circ_ARHGEF28 inhibits MST1/2 dimerization to suppress Hippo pathway to induce cisplatin resistance in ovarian cancer.

BACKGROUND: Cisplatin is integral to ovarian cancer treatment, yet resistance to this drug often results in adverse patient outcomes. The association of circular RNA (circRNA) with cisplatin resistance in ovarian cancer has been observed, but the mechanisms governing this relationship require further elucidation. METHODS: High-throughput sequencing was utilized to profile circRNA expression in cisplatin-resistant ovarian cancer cells. Gain-and-loss-of-function experiments assessed the impact on cisplatin sensitivity, both in vitro and in vivo. Fluorescence in situ hybridization was conducted to determine the cellular distribution of circRNAs, and RNA pulldown and immunoprecipitation experiments were performed to identify associated binding proteins. RESULTS: The study revealed that circ_ARHGEF28 is overexpressed in certain cisplatin-resistant ovarian cancer tissues and cell lines, and is associated with reduced progression-free survival in patients. It was observed that circ_ARHGEF28 contributes to cisplatin resistance in ovarian cancer models, both in vitro and in vivo. Importantly, circ_ARHGEF28 was found to interact directly with MST1/2, inhibiting the SARAH coiled-coil binding domains and consequently deactivating the Hippo pathway. CONCLUSION: This investigation identifies circ_ARHGEF28 as a novel circRNA that contributes to cisplatin resistance in ovarian cancer by suppressing the Hippo pathway. Therapeutic strategies targeting circ_ARHGEF28 may offer a potential avenue to mitigate cisplatin resistance in ovarian cancer treatment.

A privacy-preserving platform oriented medical healthcare and its application in identifying patients with candidemia.

Federated learning (FL) has emerged as a significant method for developing machine learning models across multiple devices without centralized data collection. Candidemia, a critical but rare disease in ICUs, poses challenges in early detection and treatment. The goal of this study is to develop a privacy-preserving federated learning framework for predicting candidemia in ICU patients. This approach aims to enhance the accuracy of antifungal drug prescriptions and patient outcomes. This study involved the creation of four predictive FL models for candidemia using data from ICU patients across three hospitals in China. The models were designed to prioritize patient privacy while aggregating learnings across different sites. A unique ensemble feature selection strategy was implemented, combining the strengths of XGBoost's feature importance and statistical test p values. This strategy aimed to optimize the selection of relevant features for accurate predictions. The federated learning models demonstrated significant improvements over locally trained models, with a 9% increase in the area under the curve (AUC) and a 24% rise in true positive ratio (TPR). Notably, the FL models excelled in the combined TPR + TNR metric, which is critical for feature selection in candidemia prediction. The ensemble feature selection method proved more efficient than previous approaches, achieving comparable performance. The study successfully developed a set of federated learning models that significantly enhance the prediction of candidemia in ICU patients. By leveraging a novel feature selection method and maintaining patient privacy, the models provide a robust framework for improved clinical decision-making in the treatment of candidemia.

Unlocking Efficient Hydrogen Production: Nucleophilic Oxidation Reactions Coupled with Water Splitting.

Electrocatalytic water splitting driven by sustainable energy is a clean and promising water-chemical fuel conversion technology for the production of high-purity green hydrogen. However, the sluggish kinetics of anodic oxygen evolution reaction (OER) pose challenges for large-scale hydrogen production, limiting its efficiency and safety. Recently, the anodic OER has been replaced by a nucleophilic oxidation reaction (NOR) with biomass as the substrate and coupled with a hydrogen evolution reaction (HER), which has attracted great interest. Anode NOR offers faster kinetics, generates high-value products, and reduces energy consumption. By coupling NOR with hydrogen evolution reaction, hydrogen production efficiency can be enhanced while yielding high-value oxidation products or degrading pollutants. Therefore, NOR-coupled HER hydrogen production is another new green electrolytic hydrogen production strategy after electrolytic water hydrogen production, which is of great significance for realizing sustainable energy development and global decarbonization. This review explores the potential of nucleophilic oxidation reactions as an alternative to OER and delves into NOR mechanisms, guiding future research in NOR-coupled hydrogen production. It assesses different NOR-coupled production methods, analyzing reaction pathways and catalyst effects. Furthermore, it evaluates the role of electrolyzers in industrialized NOR-coupled hydrogen production and discusses future prospects and challenges. This comprehensive review aims to advance efficient and economical large-scale hydrogen production.

[Risk factors for bronchopulmonary dysplasia in twin preterm infants: a multicenter study]. / åŒèƒŽæ—©äº§å„¿æ”¯æ°”ç®¡è‚ºå‘è‚²ä¸è‰¯å±é™©å› ç´ åˆ†æžï¼šä¸€é¡¹å¤šä¸­å¿ƒç ”ç©¶.

OBJECTIVES: To investigate the risk factors for bronchopulmonary dysplasia (BPD) in twin preterm infants with a gestational age of <34 weeks, and to provide a basis for early identification of BPD in twin preterm infants in clinical practice. METHODS: A retrospective analysis was performed for the twin preterm infants with a gestational age of <34 weeks who were admitted to 22 hospitals nationwide from January 2018 to December 2020. According to their conditions, they were divided into group A (both twins had BPD), group B (only one twin had BPD), and group C (neither twin had BPD). The risk factors for BPD in twin preterm infants were analyzed. Further analysis was conducted on group B to investigate the postnatal risk factors for BPD within twins. RESULTS: A total of 904 pairs of twins with a gestational age of <34 weeks were included in this study. The multivariate logistic regression analysis showed that compared with group C, birth weight discordance of >25% between the twins was an independent risk factor for BPD in one of the twins (OR=3.370, 95%CI: 1.500-7.568, P<0.05), and high gestational age at birth was a protective factor against BPD (P<0.05). The conditional logistic regression analysis of group B showed that small-for-gestational-age (SGA) birth was an independent risk factor for BPD in individual twins (OR=5.017, 95%CI: 1.040-24.190, P<0.05). CONCLUSIONS: The development of BPD in twin preterm infants is associated with gestational age, birth weight discordance between the twins, and SGA birth.

Autism spectrum disorders detection based on multi-task transformer neural network.

Autism Spectrum Disorders (ASD) are neurodevelopmental disorders that cause people difficulties in social interaction and communication. Identifying ASD patients based on resting-state functional magnetic resonance imaging (rs-fMRI) data is a promising diagnostic tool, but challenging due to the complex and unclear etiology of autism. And it is difficult to effectively identify ASD patients with a single data source (single task). Therefore, to address this challenge, we propose a novel multi-task learning framework for ASD identification based on rs-fMRI data, which can leverage useful information from multiple related tasks to improve the generalization performance of the model. Meanwhile, we adopt an attention mechanism to extract ASD-related features from each rs-fMRI dataset, which can enhance the feature representation and interpretability of the model. The results show that our method outperforms state-of-the-art methods in terms of accuracy, sensitivity and specificity. This work provides a new perspective and solution for ASD identification based on rs-fMRI data using multi-task learning. It also demonstrates the potential and value of machine learning for advancing neuroscience research and clinical practice.