Search

BackgroundThis study applied machine learning (ML) algorithms to construct a model for predicting EN initiation for patients in the intensive care unit (ICU) and identifying populations in need of EN at an early stage.MethodsThis study collected patient information from the Medical Information Mart for Intensive Care IV database. All patients enrolled were split randomly into a training set and a validation set. Six ML models were established to evaluate the initiation of EN, and the best model was determined according to the area under curve (AUC) and accuracy. The best model was interpreted using the Local Interpretable Model-Agnostic Explanations (LIME) algorithm and SHapley Additive exPlanation (SHAP) values.ResultsA total of 53,150 patients participated in the study. They were divided into a training set (42,520, 80%) and a validation set (10,630, 20%). In the validation set, XGBoost had the optimal prediction performance with an AUC of 0.895. The SHAP values revealed that sepsis, sequential organ failure assessment score, and acute kidney injury were the three most important factors affecting EN initiation. The individualized forecasts were displayed using the LIME algorithm.ConclusionThe XGBoost model was established and validated for early prediction of EN initiation in ICU patients.

This paper presents new zircon U-Pb geochronological, Hf isotopic and whole-rock geochemical data for the granitic plutons in the Xing'an Massif, Northeast China, to constrain the Late Mesozoic tectonic evolution of the Mongol-Okhotsk Ocean and the Paleo-Pacific Ocean. The zircon U-Pb ages indicate that the granitoids emplaced during the Late Jurassic-Early Cretaceous. The granodiorites show an adakitic affinity with high Sr/Y ratios and low Yb (< 1.30 μg/g) contents. The monzogranites exhibit high SiO2, low MgO contents, enrichment in LILEs (Rb, K, and Th), and depletion in HSFEs (Ta, Nb, Zr, P, and Ti). Petrological and geochemical features of these monzogranites suggest that they are highly fractionated I-type granitoids. In addition, the zircon εHf(t) values and two-stage model ages (TDM2) are in the range of +2.6 to +8.1 and 669-1011 Ma, respectively, indicating that primary magma was generated by partial melting of juvenile lower-crustal materials, and there was a significant crustal growth in the Phanerozoic in the Northeast China. Combined with the coeval granitoids widely exposed in the Xing'an Massif, we conclude that the Late Jurassic magma in Northeast China was generated in an extensional setting related to the closure of the Mongol-Okhotsk Ocean, but the Early Cretaceous magma was related to the subduction of the Paleo-Pacific Plate. [ABSTRACT FROM AUTHOR]

Electroencephalography (EEG) is considered the output of a brain and it is a bioelectrical signal with multiscale and nonlinear properties. Motor Imagery EEG (MI-EEG) not only has a close correlation with the human imagination and movement intention but also contains a large amount of physiological or disease information. As a result, it has been fully studied in the field of rehabilitation. To correctly interpret and accurately extract the features of MI-EEG signals, many nonlinear dynamic methods based on entropy, such as Approximate Entropy (ApEn), Sample Entropy (SampEn), Fuzzy Entropy (FE), and Permutation Entropy (PE), have been proposed and exploited continuously in recent years. However, these entropy-based methods can only measure the complexity of MI-EEG based on a single scale and therefore fail to account for the multiscale property inherent in MI-EEG. To solve this problem, Multiscale Sample Entropy (MSE), Multiscale Permutation Entropy (MPE), and Multiscale Fuzzy Entropy (MFE) are developed by introducing scale factor. However, MFE has not been widely used in analysis of MI-EEG, and the same parameter values are employed when the MFE method is used to calculate the fuzzy entropy values on multiple scales. Actually, each coarse-grained MI-EEG carries the characteristic information of the original signal on different scale factors. It is necessary to optimize MFE parameters to discover more feature information. In this paper, the parameters of MFE are optimized independently for each scale factor, and the improved MFE (IMFE) is applied to the feature extraction of MI-EEG. Based on the event-related desynchronization (ERD)/event-related synchronization (ERS) phenomenon, IMFE features from multi channels are fused organically to construct the feature vector. Experiments are conducted on a public dataset by using Support Vector Machine (SVM) as a classifier. The experiment results of 10-fold cross-validation show that the proposed method yields relatively high classification accuracy compared with other entropy-based and classical time–frequency–space feature extraction methods. The t-test is used to prove the correctness of the improved MFE.