RESUMO
BACKGROUND: Acute kidney injury (AKI) is not only a complication but also a serious threat to patients with cerebral infarction (CI). This study aimed to explore the application of interpretable machine learning algorithms in predicting AKI in patients with cerebral infarction. METHODS: The study included 3920 patients with CI admitted to the Intensive Care Unit and Emergency Medicine of the Central Hospital of Lishui City, Zhejiang Province. Nine machine learning techniques, including XGBoost, logistics, LightGBM, random forest (RF), AdaBoost, GaussianNB (GNB), Multi-Layer Perceptron (MLP), support vector machine (SVM), and k-nearest neighbors (KNN) classification, were used to develop a predictive model for AKI in these patients. SHapley Additive exPlanations (SHAP) analysis provided visual explanations for each patient. Finally, model effectiveness was assessed using metrics such as average precision (AP), sensitivity, specificity, accuracy, F1 score, precision-recall (PR) curve, calibration plot, and decision curve analysis (DCA). RESULTS: The XGBoost model performed better in the internal validation set and the external validation set, with an AUC of 0.940 and 0.887, respectively. The five most important variables in the model were, in order, glomerular filtration rate, low-density lipoprotein, total cholesterol, hemiplegia and serum kalium. CONCLUSION: This study demonstrates the potential of interpretable machine learning algorithms in predicting CI patients with AKI.
Assuntos
Injúria Renal Aguda , Infarto Cerebral , Unidades de Terapia Intensiva , Aprendizado de Máquina , Valor Preditivo dos Testes , Humanos , Injúria Renal Aguda/diagnóstico , Injúria Renal Aguda/sangue , Injúria Renal Aguda/terapia , Masculino , Feminino , Idoso , Pessoa de Meia-Idade , Infarto Cerebral/diagnóstico , Infarto Cerebral/etiologia , Fatores de Risco , Medição de Risco , China/epidemiologia , Prognóstico , Reprodutibilidade dos Testes , Idoso de 80 Anos ou mais , Técnicas de Apoio para a Decisão , Estudos Retrospectivos , Diagnóstico por ComputadorRESUMO
Introduction: Acute liver injury (ALI) is a common complication of sepsis and is associated with adverse clinical outcomes. We aimed to develop a model to predict the risk of ALI in patients with sepsis after hospitalization. Methods: Medical records of 3196 septic patients treated at the Lishui Central Hospital in Zhejiang Province from January 2015 to May 2023 were selected. Cohort 1 was divided into ALI and non-ALI groups for model training and internal validation. The initial laboratory test results of the study subjects were used as features for machine learning (ML), and models built using nine different ML algorithms were compared to select the best algorithm and model. The predictive performance of model stacking methods was then explored. The best model was externally validated in Cohort 2. Results: In Cohort 1, LightGBM demonstrated good stability and predictive performance with an area under the curve (AUC) of 0.841. The top five most important variables in the model were diabetes, congestive heart failure, prothrombin time, heart rate, and platelet count. The LightGBM model showed stable and good ALI risk prediction ability in the external validation of Cohort 2 with an AUC of 0.815. Furthermore, an online prediction website was developed to assist healthcare professionals in applying this model more effectively. Conclusions: The Light GBM model can predict the risk of ALI in patients with sepsis after hospitalization.
RESUMO
The research and development of aerogel-based microwave absorbing materials with strong electromagnetic (EM) wave response is an emerging research topic in the EM wave absorption field. In order to implement light microwave absorbers with a broad bandwidth, freeze drying assisted with in situ thermally structure-directing techniques was applied to fabricate composite aerogels with orientation design. Thanks to the integration of pore structure regulation and conductive network construction, the as-prepared aerogel absorbers exhibit a tunable EM response covering a broad frequency range. In detail, the maximum reflection loss (RL) value of the CR-3 aerogel reaches -50.8 dB at 2.2 mm and its maximum effective absorption bandwidth reaches 5.4 GHz at 2.0 mm, which is in accordance with the numerical simulation results of the radar cross section (RCS), where the optimum RCS reduction of 21.4 dB m2 appears for the CR-3 aerogel when the detection theta was set as 0°. In all, this work paves the way for the exploration of high-efficiency aerogel absorbers by balancing the evolution of the pore structure and conductive connection at the same time.
RESUMO
Metamaterials are usually designed using biomimetic technology based on natural biological characteristics or topology optimization based on prior knowledge. Although satisfactory results can be achieved to a certain extent, there are still many performance limitations. For overcoming the above limitations, this paper proposes a rapid metamaterials design method based on the generation of random topological patterns. This method realizes the combined big data simulation and structure optimization of structure-electromagnetic properties, which makes up for the shortcomings of traditional design methods. The electromagnetic properties of the proposed metamaterials are verified by experiments. The reflection coefficient of the designed absorbing metamaterial unit is all lower than -15 dB over 12-16 GHz. Compared with the metal floor, the radar cross section (RCS) of the designed metamaterial is reduced by a minimum of 14.5 dB and a maximum of 27.6 dB over the operating band. The performance parameters of metamaterial obtained based on the random topology design method are consistent with the simulation design results, which further verifies the reliability of the algorithm in this paper.
RESUMO
A series of manganese(I) carbonyl complexes bearing structurally related NN- and NNN-chelating ligands have been synthesized and assessed as catalysts for transfer hydrogenation (TH). Notably, the NN-systems based on N-R functionalized 5,6,7,8-tetrahydroquinoline-8-amines, proved the most effective in the manganese-promoted conversion of acetophenone to 1-phenylethanol. In particular, the N-isopropyl derivative, Mn1, when conducted in combination with t-BuONa, was the standout performer mediating not only the reduction of acetophenone but also a range of carbonyl substrates including (hetero)aromatic-, aliphatic- and cycloalkyl-containing ketones and aldehydes with especially high values of TON (up to 17 200; TOF of 3550 h-1). These findings, obtained through a systematic variation of the N-R group of the NN ligand, are consistent with an outer-sphere mechanism for the hydrogen transfer. As a more general point, this Mn-based catalytic TH protocol offers an attractive and sustainable alternative for producing alcoholic products from carbonyl substrates.