Search

As the system with the highest solar hydrogen conversion efficiency, the further improvement of photovoltaic-photoelectrocatalysis (PV-PEC) efficiency is dependent on the light transmittance, activity and stability of the photoanode. Here, a highly permeable BiVO 4 thin film was fabricated through controlled magnetron sputtering. The thickness of the thin film was merely 345 nm, while its transmittance in the visible light range reached as high as 70%. Moreover, large-scale production of such films could be achieved. By incorporating NiFeO x oxygen extraction cocatalyst, under AM 1.5G illumination, the photocurrent density at 1.23 V on reversible hydrogen electrode (RHE) increased to 4.2 mA cm−2 and the incident photon-current conversion efficiency reached 70%. A PV-PEC system was designed to efficiently harness sunlight by integrating the photoanode with a solar photovoltaic panel. Upon connection with the solar photovoltaic panel (with an efficiency of only 7%), spontaneous water decomposition could be achieved, resulting in a solar-to-hydrogen conversion efficiency (η STH) of 4.2% for the series system. This study contributes to the realization of cost-effective and highly efficient PV-PEC hydrogen production. The PV-PEC series system with BiVO 4 thin film photoanode prepared by magnetron sputtering method improves the overall utilization rate of sunlight and realizes spontaneous water decomposition. [Display omitted] • BiVO 4 films with large area and high transmittance can be prepared by low rate controlled magnetron sputtering. • The PEC performance of films was improved by optimizing the magnetron sputtering process and supporting catalyst. • Photovoltaic and photoanode in PV-PEC cells can be irradiated in series to improve the utilization rate of sunlight. [ABSTRACT FROM AUTHOR]

Objective: To create a comprehensive map of strategic lesion network localizations for neurological deficits, and identify prognostic neuroimaging biomarkers to facilitate the early detection of patients with a high risk of poor functional outcomes in acute ischemic stroke (AIS). Methods: In a large‐scale multicenter study of 7,807 patients with AIS, we performed voxel‐based lesion‐symptom mapping, functional disconnection mapping (FDC), and structural disconnection mapping (SDC) to identify distinct lesion and network localizations for National Institutes of Health Stroke Scale (NIHSS) score. Impact scores were calculated based on the odds ratios or t‐values of voxels from voxel‐based lesion‐symptom mapping, FDC, and SDC results. Ordinal regression models were used to investigate the predictive value of the impact scores on functional outcome (defined as the modified Rankin score at 3 months). Results: We constructed lesion, FDC, and SDC maps for each item of the NIHSS score, which provided insights into the neuroanatomical substrate and network localization of neurological function deficits after AIS. The lesion impact score of limb ataxia, the SDC impact score of limb deficit, and FDC impact score of sensation and dysarthria were significantly associated with modified Rankin Scale at 3 months. Adding the SDC impact score, FDC impact score, and lesion impact score to the NIHSS total score improved the performance in predicting functional outcomes, as compared with using the NIHSS score alone. Interpretation: We constructed comprehensive maps of strategic lesion network localizations for neurological deficits that were predictive of functional outcomes in AIS. These results may provide specifically localized targets for future neuromodulation therapies. ANN NEUROL 2023;94:572–584 [ABSTRACT FROM AUTHOR]

[Display omitted] • The regulation of trace Cu+ ions on the structures and band of Cu x In 0.25 ZnS y solid solution are studied. • The mechanism of energy band bending and matching between Cu x In 0.25 ZnS y and cocatalysts are revealed. • The optimized Cu x In 0.25 ZnS y @Rh photocatalyst exhibits a high catalytic HER activity under the visible light irradiation. As an attractive semiconductor photocatalyst, (CuInS 2) x -(ZnS) y has been intensively studied in photocatalysis, due to its unique layered structure and stability. Here, we synthesized a series of Cu x In 0.25 ZnS y photocatalysts with different trace Cu+-dominated ratios. The results show that doping with Cu+ ions leads to an increase in the valence state of In and the formation of a distorted S structure, simultaneously inducing a decrease in the semiconductor bandgap. When the doping amount of Cu+ ions is 0.04 atomic ratio to Zn, the optimized Cu 0.04 In 0.25 ZnS y photocatalyst with a bandgap of 2.16 eV shows the highest catalytic hydrogen evolution activity (191.4 μmol.h−1). Subsequently, among the common cocatalysts, Rh loaded Cu 0.04 In 0.25 ZnS y gives the highest activity of 1189.8 μmol·h−1, corresponding to an apparent quantum efficiency of 49.11 % at 420 nm. Moreover, the internal mechanism of photogenerated carrier transfer between semiconductors and different cocatalysts is analyzed by the band bending phenomenon. [ABSTRACT FROM AUTHOR]

Background and purpose: Acute brainstem infarctions can lead to serious functional impairments. We aimed to predict functional outcomes in patients with acute brainstem infarction using deep neuroimaging features extracted by convolutional neural networks (CNNs). Methods: This nationwide multicenter stroke registry study included 1482 patients with acute brainstem infarction. We applied CNNs to automatically extract deep neuroimaging features from diffusion‐weighted imaging. Deep learning models based on clinical features, laboratory features, conventional imaging features (infarct volume, number of infarctions), and deep neuroimaging features were trained to predict functional outcomes at 3 months poststroke. Unfavorable outcome was defined as modified Rankin Scale score of 3 or higher at 3 months. The models were evaluated by comparing the area under the receiver operating characteristic curve (AUC). Results: A model based solely on 14 deep neuroimaging features from CNNs achieved an extremely high AUC of 0.975 (95% confidence interval [CI] = 0.934–0.997) and significantly outperformed the model combining clinical, laboratory, and conventional imaging features (0.772, 95% CI = 0.691–0.847, p < 0.001) in prediction of functional outcomes. The deep neuroimaging model also demonstrated significant improvement over traditional prognostic scores. In an interpretability analysis, the deep neuroimaging features displayed a significant correlation with age, National Institutes of Health Stroke Scale score, infarct volume, and inflammation factors. Conclusions: Deep learning models can successfully extract objective neuroimaging features from the routine radiological data in an automatic manner and aid in predicting the functional outcomes in patients with brainstem infarction at 3 months with very high accuracy. [ABSTRACT FROM AUTHOR]

The paper introduces a robust methodology for evaluating complex electricity system efficiency, making several contributions. It systematically decomposes the electric power system into various processes, including power generation, transmission, distribution, and power consumption, with parallel sub-processes for production and domestic power consumption. Employing data envelopment analysis (DEA), the paper develops a network directional slacks-based (DSB) model to evaluate the whole system and the internal processes. The model is theoretically proven to be feasible, providing efficiency scores and identifying inefficiencies in the system and its processes. A noteworthy addition is the introduction of the DSB green Luenberger productivity index and its decomposition, offering an analysis of productivity changes in electricity systems. Empirical analysis covering 30 Chinese provinces from 2013 to 2019 uncovers the performance variations among provinces and processes, pointing out inefficiencies and changes in technology and efficiency over time. The empirical findings aim to support efforts in enhancing the overall efficiency of China's interprovincial power system. • We decompose the complex electric power system into four processes with dependence. • We build a network directional slacks-based (DSB) model for the complex power system. • We give a DSB green Luenberger productivity index (DPI) for our network model. • The efficiency score and DPI of the system are decomposed into those of processes. • We apply the model and indexes to electric power systems in Chinese provinces. [ABSTRACT FROM AUTHOR]

The structural, electronic, and optical properties of the AgnCun (n = 7–14) clusters with the equal number of the Ag and Cu atoms are systematically investigated by using the density functional theory method. The results indicate that as n is 9, the structures of the AgnCun clusters begin to form the icosahedral core, and most Ag–Cu clusters with icosahedral cores have high stability. The Ag8Cu8 cluster has the highest mixing degree. The Ag10Cu10 cluster has relatively high energetic and electronic stability. For the density of states at the Fermi level, the Ag7Cu7, Ag8Cu8, Ag9Cu9, Ag10Cu10, and Ag14Cu14 clusters are mainly contributed by Cu atoms, while the Ag11Cu11, Ag12Cu12, and Ag13Cu13 clusters are mainly contributed by Ag atoms. Interestingly, the Cu16-II cluster is magnetic. The absorption spectra of the AgnCun clusters are red-shift comparing with the pure Ag2n clusters, but they are blue-shift comparing with the pure Cu2n clusters. The absorption properties of the AgnCun clusters in the visible region are better than the pure clusters. [ABSTRACT FROM AUTHOR]

Ventilator-induced lung injury (VILI) is the most common complication associated with mechanical ventilation. Electroacupuncture (EA) has shown potent anti-inflammatory effects. This study aimed to investigate the effects of EA on VILI and explore the underlying mechanisms. Male C57BL/6 mice were subjected to high tidal volume ventilation to induce VILI. Prior to mechanical ventilation, mice received treatment with EA, nonacupoint EA, or EA combined with zinc protoporphyrin. EA treatment significantly improved oxygenation, as indicated by increased PaO 2 levels in VILI mice. Moreover, EA reduced lung injury score, lung wet/dry weight ratio, and protein concentration in bronchoalveolar lavage fluid. EA also decreased the expression of pro-inflammatory cytokines including interleukin (IL)-1β, IL-6, tumor necrosis factor-α, IL-18, chemokine keratinocyte chemoattractant, macrophage inflammatory protein 2, and malondialdehyde. Furthermore, EA increased the activities of antioxidant enzymes superoxide dismutase, catalase, and glutathione peroxidase in VILI mice. At the molecular level, EA upregulated the expression of Nrf2 (nucleus) and heme oxygenase −1, while down-regulating the expression of p-NF-κB p65, NLR Family Pyrin Domain Containing 3, Cleaved Caspase-1, and ASC in VILI mice. Notably, the effects of EA were reversed by zinc protoporphyrin treatment, nonacupoint EA did not affect the aforementioned indicators of VILI. EA alleviates VILI by inhibiting the NLR Family Pyrin Domain Containing three inflammasome through activation of the Nrf2/HO-1 pathway. [ABSTRACT FROM AUTHOR]

Perioperative neurocognitive disorders (PND) increases postoperative dementia and mortality in patients and has no effective treatment. Although the detailed pathogenesis of PND is still elusive, a large amount of evidence suggests that damaged mitochondria may play an important role in the pathogenesis of PND. A healthy mitochondrial pool not only provides energy for neuronal metabolism but also maintains neuronal activity through other mitochondrial functions. Therefore, exploring the abnormal mitochondrial function in PND is beneficial for finding promising therapeutic targets for this disease. This article summarizes the research advances of mitochondrial energy metabolism disorder, inflammatory response and oxidative stress, mitochondrial quality control, mitochondria-associated endoplasmic reticulum membranes, and cell death in the pathogenesis of PND, and briefly describes the application of mitochondria-targeted therapies in PND. [ABSTRACT FROM AUTHOR]

Background: Medical imaging reports play an important role in communication of diagnostic information between radiologists and clinicians. Head magnetic resonance imaging (MRI) reports can provide evidence that is widely used in the diagnosis and treatment of ischaemic stroke. The high-signal regions of diffusion-weighted imaging (DWI) images in MRI reports are key evidence. Correctly identifying high-signal regions of DWI images is helpful for the treatment of ischaemic stroke patients. Since most of the multiple signals recorded in head MRI reports appear in the same part, it is challenging to identify high-signal regions of DWI images from MRI reports.Methods: We developed a deep learning model to automatically identify high-signal regions of DWI images from head MRI reports. We proposed a fine-grained entity typing model based on machine reading comprehension that transformed the traditional two-step fine-grained entity typing task into a question-answering task.Results: To prove the validity of the model proposed, we compared it with the fine-grained entity typing model, of which the F1 measure was 5.9% and 3.2% higher than the F1 measures of the models based on LSTM and BERT, respectively.Conclusion: In this study, we explore the automatic identification of high-signal regions of DWI images from the description part of a head MRI report. We transformed the identification of high-signal regions of DWI images to an FET task and proposed an MRC-FET model. Compared with the traditional two-step FET method, the model we proposed not only simplifies the task but also has better performance. The comparable result shows that the work in this study can contribute to improving the clinical decision support system. [ABSTRACT FROM AUTHOR]

The balance centered at Cr3+, La3+ and Al3+ was investigated and the stabilizing mechanism of Cr3+ was proposed during crystal remolding. The crystallinity of bulk STO perovskite is improved by the molten-salt method, and the exposure of new non-equivalent facets is more conducive to internal charge separation and migration. The above development is helpful for the overall water splitting under visible light. [Display omitted] • Designed single-crystalline La/(Cr, Al) co-doped SrTiO 3 was successfully prepared. • Introduced Cr3+ for visible-light excitation is fixed and stabilized by La3+ and Al3+. • Ti3+ defect concentration is effectively suppressed after crystal remolding. • The designed SrTiO 3 material performed well in visible-light-driven water splitting. Achieving efficient solar-to-hydrogen conversion through visible-light-driven overall water splitting poses a significant challenge for particulate photocatalysts. Although bandgap engineering can be realized by ion doping, the introduced defects simultaneously enhance the recombination of electron-hole pairs. Herein, La/(Cr, Al) co-doped SrTiO 3 was remolded via the ion implantation flux method. The La, Cr and Al achieve valence state equilibrium upon stabilizing the crystal lattice, resulting in visible-light absorption. The optimized La/(Cr, Al)-SrTiO 3 photocatalyst exhibits an overall water-splitting reaction under visible light, yielding 73.2 μmol·h−1 for H 2 and 35.0 μmol·h−1 for O 2 with an apparent quantum yield of 1.44 % at 420 nm. Based on theoretical calculation, we proposed a La, Al-cooperated mechanism that a little Cr3+ was fixed by the surrounding Al3+, and the valence state was stabilized by La3+ during the crystal remolding, thereby giving the guidance for designing visible-light materials applied in overall water splitting. [ABSTRACT FROM AUTHOR]

STUDY QUESTION Do women have worse pregnancy and neonatal outcomes of IVF/ICSI–fresh embryo transfer (ET) after conservative treatment of atypical hyperplasia (AH)? SUMMARY ANSWER AH has no impact on live birth but is associated with increased risks of pregnancy loss and preterm delivery (PTD). WHAT IS KNOWN ALREADY AH is a precancerous lesion of endometrial cancer. Several recognized AH risk factors include nulliparity, increased body mass index, ovulation disorders, diabetes mellitus, and others. As such, patients are suggested to attempt conception upon achieving AH regression. Recently, successful pregnancies with IVF/ICSI have been increasingly reported. STUDY DESIGN, SIZE, DURATION Forty-two patients with AH regression and 18 700 women with no evidence of endometrial abnormality, who underwent their first autologous oocytes' retrieval and fresh ET cycles of IVF/ICSI in the Center for Reproductive Medicine, Shandong University, from May 2008 to July 2021, were retrospectively enrolled. PARTICIPANTS/MATERIALS, SETTING, METHODS First, 42 AH patients were propensity score matched with control women (n = 168) at a 1:4 ratio. Reproductive outcomes and maternal/neonatal complications were compared between the matched pairs. Binary logistic regression analyses were conducted to assess odds ratios (ORs) of AH for live birth, pregnancy loss, and PTD from AH women and all 18 700 eligible controls. MAIN RESULT AND THE ROLE OF CHANCE Patients with AH achieved a numerically lower live birth rate (LBR) as compared to the matched controls, but without significant difference (26% versus 37%, P  = 0.192). However, compared with the matched controls, AH patients showed significantly higher rates of pregnancy loss (52% versus 21%, P  = 0.003) and PTD (45% versus 16%, P  = 0.041). Further analyses revealed a statistically significantly increased rate of late pregnancy loss (17% versus 3%, P  = 0.023), but not early miscarriage (35% versus 18%, P  = 0.086), in the AH group. Furthermore, after correcting for potential confounders, the likelihood of a live birth in AH patients narrowly failed to be statistically significantly different from controls (adjusted OR [aOR]: 0.51, 95% CI: 0.25–1.04, P  = 0.064). Nonetheless, the logistic regression reconfirmed that AH was an independent risk factor for pregnancy loss (aOR: 3.62, 95% CI: 1.55–8.46, P  = 0.003), late pregnancy loss (aOR: 9.33, 95% CI: 3.00–29.02, P  < 0.001), and PTD (aOR: 5.70, 95% CI: 1.45–22.38, P  = 0.013). LIMITATIONS, REASONS FOR CAUTION Selection bias was an inherent drawback of this study. First, because of the low AH prevalence among women receiving IVF/ICSI treatment, and consequently, limited sample size, the relationship between AH with LBR and adverse complications might be concealed and underestimated. Hence, the results should be interpreted cautiously. Similarly, the impacts of diverse clinical features of AH patients on the pregnancy outcomes need further studies in a larger population. Second, although most data used in this study were obtained by reviewing the medical records, missing data did exist and so did the recall bias. Third, although the propensity score matching and multivariable logistic models were performed collectively in order to minimize potential confounders between AH and controls, the intrinsic disadvantages of the retrospective nature of this study could not be avoided completely, and additional confirmation bias might be induced with reduplication of statistical analyses. WIDER IMPLICATION OF THE FINDINGS Our results highlight the necessity of adequate counseling and intensive pregnancy monitoring for AH individuals and their families. STUDY FUNDING/COMPETING INTEREST(S) This work was supported by grants from the National Key Research & Developmental Program of China (2022YFC2703800), the Natural Science Foundation of Shandong Province (ZR2022MH009), and Projects of Medical and Health Technology Development Program in Shandong Province (202005010520, 202005010523). There are no conflicts of interest to declare. TRIAL REGISTRATION NUMBER N/A. [ABSTRACT FROM AUTHOR]

Neurons are highly dependent on mitochondrial ATP production and Ca2+ buffering. Neurons have unique compartmentalized anatomy and energy requirements, and each compartment requires continuously renewed mitochondria to maintain neuronal survival and activity. Peroxisome proliferator-activated receptor-gamma coactivator-1α (PGC-1α) is a key factor in the regulation of mitochondrial biogenesis. It is widely accepted that mitochondria are synthesized in the cell body and transported via axons to the distal end. However, axonal mitochondrial biogenesis is necessary to maintain axonal bioenergy supply and mitochondrial density due to limitations in mitochondrial axonal transport rate and mitochondrial protein lifespan. In addition, impaired mitochondrial biogenesis leading to inadequate energy supply and neuronal damage has been observed in neurological disorders. In this review, we focus on the sites where mitochondrial biogenesis occurs in neurons and the mechanisms by which it maintains axonal mitochondrial density. Finally, we summarize several neurological disorders in which mitochondrial biogenesis is affected. [ABSTRACT FROM AUTHOR]

• Bone marrow-derived mesenchymal stem cells (BMSCs) transplantation may retard renal fibrosis after acute kidney injury. • BMSCs transplantation may diminish pericyte-myofibroblast transition. • BMSCs transplantation may diminish extracellular matrix augment. • BMSCs transplantation may regulate transforming growth factor β/Smad signaling pathway. Renal fibrosis is a common chronic outcome of acute kidney injury (AKI). Pericyte-myofibroblasts transition and production of abundant extracellular matrix are the important pathologic basis. This study investigated the effect of bone marrow-derived mesenchymal stem cells (BMSCs) transplantation on the AKI kidney fibrosis and the possible mechanisms. By constructing the animal and cell model of AKI pericyte injury, the therapeutic effect of BMSCs on pericyte-myofibroblasts transition was detected. The production and accumulation of extracellular matrix, including collagen I, collagen III, and fibronectin were also tested. The mechanism was revealed by means of analysis of signal pathway. After AKI insult, many myofibroblasts emerged in the renal interstitium together with a large amount of extracellular matrix components. The BMSCs transplantation significantly decreased the number of myofibroblasts trans-differentiated from pericytes in the AKI model. The changes of vascular endothelial growth factor subtypes and Ang-I/AngII secreted by pericytes were also significantly reduced after BMSCs co-culture. At the same time, extracellular matrix components, including collagen I, collagen III, and fibronectin, decreased significantly. Transplantation treatment alleviated the fibrosis score. The transforming growth factor β (TGF-β) concentration decreased as well as the levels of Smad2/3 and p-Smad2/3 with the presence of BMSCs therapy. Bone marrow-derived mesenchymal stem cells transplantation diminished pericyte-myofibroblast transition and extracellular matrix augment after AKI by regulating the TGF-β/Smad2/3 signaling pathway. It may be used as a novel therapeutic method for retarding renal fibrosis, which is worthy of further study. [ABSTRACT FROM AUTHOR]

Purpose: We develop a new risk score to predict patients with stroke-associated pneumonia (SAP) who have an acute intracranial hemorrhage (ICH).Method: We applied logistic regression to develop a new risk score called ICH-LR2S2. It was derived from examining a dataset of 70,540 ICH patients between 2015 and 2018 from the Chinese Stroke Center Alliance (CSCA). During the training of ICH-LR2S2, patients were randomly divided into two groups - 80% for the training set and 20% for model validation. A prospective test set was developed using 12,523 patients recruited in 2019. To further verify its effectiveness, we tested ICH-LR2S2 on an external dataset of 24,860 patients from the China National Stroke Registration Management System II (CNSR II). The performance of ICH-LR2S2 was measured by the area under the receiver operating characteristic curve (AUROC).Results: The incidence of SAP in the dataset was 25.52%. A 24-point ICH-LR2S2 was developed from independent predictors, including age, modified Rankin Scale, fasting blood glucose, National Institutes of Health Stroke Scale admission score, Glasgow Coma Scale score, C-reactive protein, dysphagia, Chronic Obstructive Pulmonary Disease, and current smoking. The results showed that ICH-LR2S2 achieved an AUC = 0.749 [95% CI 0.739-0.759], which outperforms the best baseline ICH-APS (AUC = 0.704) [95% CI 0.694-0.714]. Compared with the previous ICH risk scores, ICH-LR2S2 incorporates fasting blood glucose and C-reactive protein, improving its discriminative ability. Machine learning methods such as XGboost (AUC = 0.772) [95% CI 0.762-0.782] can further improve our prediction performance. It also performed well when further validated by the external independent cohort of patients (n = 24,860), ICH-LR2S2 AUC = 0.784 [95% CI 0.774-0.794].Conclusion: ICH-LR2S2 accurately distinguishes SAP patients based on easily available clinical features. It can help identify high-risk patients in the early stages of diseases. [ABSTRACT FROM AUTHOR]

What role does programmed cell death 4 (PDCD4) play in premature ovarian insufficiency (POI)? A PDCD4 gene knockout (PDCD4–/–) mouse model was constructed, a POI mouse model was established similar to human POI with 4-vinylcyclohexene dioxide (VCD), a PDCD4 -overexpressed adenovirus was designed and the regulatory role in POI in vitro and in vivo was investigated. PDCD4 expression was significantly increased in the ovarian granulosa cells of patients with POI (P ≤ 0.002 protein and mRNA) and mice with VCD-induced POI (P < 0.001 protein expression in both mouse ovaries and granulosa cells). In POI-induced mice model, PDCD4 knockouts significantly increased anti-Müllerian hormone, oestrodiol and numbers of developing follicles, and the PI3K-AKT-Bcl2/Bax signalling pathway is involved in it. The expression and regulation of PDCD4 significantly affects the POI pathology in a mouse model. This effect is closely related to the regulation of Bcl2/Bax and the activation of the PI3K-AKT signalling pathway. [ABSTRACT FROM AUTHOR]

• A novel trigeneration process for crude helium, liquefied natural gas, and methanol production was proposed. • The application of WRU provided the feasibility of producing more methanol through a separation process from vapor streams. • Comprehensive energy, exergy, economic, and environmental analyses were performed for feasibility evaluation. • According to the results, the total energy and exergy efficiencies were obtained to be 94.38% and 93.4%. • The net present value of the designed scheme was estimated to be 364.36 M$. This study suggested a novel integrated trigeneration system for crude helium (He), liquefied natural gas, and methanol production. This proposed system consisted of a cryogenic-based helium recovery unit from natural gas, a syngas synthesis unit by hybrid reforming, a methanol synthesis unit, and a sour water recovery unit. Comprehensive energy, exergy, environmental, and economic analyses were performed to assess the system's performance. Based on the results, the crude helium, liquefied natural gas, and methanol production rates were obtained to be 4105 k g / h , 3594000 k g / h , and 707800 k g / h , respectively. These production rates provided 93.4 % and 94.38 % overall exergy and energy efficiencies. Also, over 60 % of the total exergy destruction occurred in reactors (R1-R4). The application of sour water recovery supplied 85.66 % of the reforming and methanol synthesis process. Regarding environmental analysis, 94 % of the CO 2 emission was due to the indirect mode, and the rate of the CO 2 emission was estimated to be 0.164 k g C O 2 / k g p r o d s . Finally, the economic analysis indicated that the total annual production costs were about 440.16 M$ and 0.13 U S D / k g. Therefore, the net present value of the designed scheme was estimated to be 364.36 M$. [ABSTRACT FROM AUTHOR]

Stroke is one of the leading causes of death and disability worldwide. The National Institutes of Health Stroke Scale (NIHSS) scores in electronic health records (EHRs), which quantitatively describe patients' neurological deficits in evidence-based treatment, are crucial in stroke-related clinical investigations. However, the free-text format and lack of standardization inhibit their effective use. Automatically extracting the scale scores from the clinical free text so that its potential value in real-world studies is realized has become an important goal. This study aims to develop an automated method to extract scale scores from the free text of EHRs. We propose a two-step pipeline method to identify NIHSS items and numerical scores and validate its feasibility using a freely accessible critical care database: MIMIC-III (Medical Information Mart for Intensive Care III). First, we utilize MIMIC-III to create an annotated corpus. Then, we investigate possible machine learning methods for two subtasks, NIHSS item and score recognition and item-score relation extraction. In the evaluation, we conduct both task-specific and end-to-end evaluations and compare our method with the rule-based method using precision, recall and F1 scores as evaluation metrics. We use all available discharge summaries of stroke cases in MIMIC-III. The annotated NIHSS corpus contains 312 cases, 2929 scale items, 2774 scores and 2733 relations. The results show that the best F1-score of our method was 0.9006, which was attained by combining BERT-BiLSTM-CRF and Random Forest, and it outperformed the rule-based method (F1-score = 0.8098). In the end-to-end task, our method could successfully recognize the item "1b level of consciousness questions", the score "1" and their relation "('1b level of consciousness questions', '1', 'has value')" from the sentence "1b level of consciousness questions: said name = 1", while the rule-based method could not. The two-step pipeline method we propose is an effective approach to identify NIHSS items, scores and their relations. With its help, clinical investigators can easily retrieve and access structured scale data, thereby supporting stroke-related real-world studies. • A two-step information extraction pipeline method is proposed to identify stroke scale scores from free text of EHRs. • The combination of BERT-BiLSTM-CRF and RF achieved the best performance. • The annotated corpus is available on PhysioNet for further research. • Automatically extracting NIHSS scores from clinical free text could help realize its potential value in real-world studies. [ABSTRACT FROM AUTHOR]