Predicting Antibiotic Resistance in ICUs Patients by Applying Machine Learning in Vietnam.

Introduction: Artificial Intelligence (AI) and machine learning (ML) are used extensively in HICs to detect and control antibiotic resistance (AMR) in laboratories and clinical institutions. ML is designed to predict outcome variables using an algorithm to enable "machines" to learn the "rules" from the data. ML is increasingly being applied in intensive care units to identify AMR and to assist empiric antibiotic therapy. This study aimed to evaluate the performance of ML models for predicting AMR bacteria and resistance to antibiotics in two Vietnamese hospitals. Patients and Methods: A cross-sectional study combined with retrospective was conducted from 1st January 2020 to 30th June 2022. Five models were developed to predict antibiotic resistance of bacterial infections of ICU patients. Two datasets were prepared to predict AMR bacteria and antibiotics with ML models. The performance of the prediction models was evaluated by various indicators (sensitivity, specificity, precision, accuracy, F1-score, PRC, AuROC, and NormMCC) to determine the optimal time point for data selection. Python version 3.8 was used for statistical analyses. Results: The accuracy, F1-score, AuROC, and normMMC of LightGBM, XGBoost, and Random Forest models were higher than those of other models in both datasets. In both datasets 1 and 2, accuracy, F1-score, AuROC and normMCC of the XGBoost model were the highest among five models (from 0.890 to 1.000). Only Random Forest models had specificity scores higher than 0.850. High scores of sensitivity, accuracy, precision, F1-score, and normMCC indicated that the models were making accurate predictions for datasets 1 and 2. Conclusion: XGBoost, LightGBM, and Random Forest were the best-performed machine learning models to predict antibiotic resistance of bacterial infections of ICUs patients using the patients' EMRs.

Thickness-dependent in-plane anisotropy of GaTe phonons.

Gallium Telluride (GaTe), a layered material with monoclinic crystal structure, has recently attracted a lot of attention due to its unique physical properties and potential applications for angle-resolved photonics and electronics, where optical anisotropies are important. Despite a few reports on the in-plane anisotropies of GaTe, a comprehensive understanding of them remained unsatisfactory to date. In this work, we investigated thickness-dependent in-plane anisotropies of the 13 Raman-active modes and one Raman-inactive mode of GaTe by using angle-resolved polarized Raman spectroscopy, under both parallel and perpendicular polarization configurations in the spectral range from 20 to 300 cm-1. Raman modes of GaTe revealed distinctly different thickness-dependent anisotropies in parallel polarization configuration while nearly unchanged for the perpendicular configuration. Especially, three Ag modes at 40.2 ([Formula: see text]), 152.5 ([Formula: see text]), and 283.8 ([Formula: see text]) cm-1 exhibited an evident variation in anisotropic behavior as decreasing thickness down to 9 nm. The observed anisotropies were thoroughly explained by adopting the calculated interference effect and the semiclassical complex Raman tensor analysis.

Effects of perinatal dioxin exposure on neonatal electroencephalography (EEG) activity of the quiet sleep stage in the most contaminated area from Agent Orange in Vietnam.

OBJECTIVE: To investigate the effects of perinatal dioxin exposure indicated by dioxins in breast milk on neonatal electroencephalography (EEG) power in the quiet sleep stage, and associations with neurodevelopmental outcomes at 2 years of age. STUDY DESIGN: Fifty-one mother-newborn pairs were enrolled for neonatal EEG analysis in the quiet sleep stage from a birth cohort recruited at a prefecture hospital in Bien Hoa city, Vietnam. Relative EEG power in intra-burst-intervals and high-voltage-bursts in the trace alternant pattern were computed from EEG data during the quiet sleep stage. Forty-three mother-child pairs participated in a 2-year follow-up survey to examine neurodevelopment using the Bayley-III scale and gaze behavior exhibited by fixation duration on the face of a child talking in videos. The general linear model and regression linear model were used for data analysis after adjusting for confounding factors. RESULTS: Perinatal dioxin exposure, particularly 2,3,7,8-tetrachlorodibenzo-p-dioxin (TCDD) exposure, influenced relative EEG power values mainly in the intra-burst-interval part of the trace alternant pattern in the quiet sleep stage. In intra-burst-intervals, decreased frontal delta power and increased frontal and parietal alpha power values in the left hemisphere and temporal beta power values in the right hemisphere were associated with increased TCDD exposure, with significant dose-response relationships. Almost none of the relative power values in these brain regions were associated with Bayley III scores, but relative delta power values were significantly associated with face fixation duration in left frontal and parietal regions at 2 years of age. CONCLUSION: Perinatal dioxin exposure influences neuronal activity in the quiet sleep stage, leading to poor communication ability indicated by gaze behavior in early childhood.

Effects of perinatal dioxin exposure on learning abilities of 8-year-old children in Vietnam.

BACKGROUND: We have followed a birth cohort from 2008 to 2009 near a dioxin-contaminated area of Da Nang, Vietnam, and investigated the effects of perinatal dioxin exposure on neurodevelopment from infancy to pre-school age. The present study aimed to investigate the effects of perinatal dioxin exposure on the learning abilities of the elementary-school children from the Da Nang birth cohort. METHODS: From 241 mother-infant pairs recruited at baseline (134 boys and 107 girls), 185 (76.8%) participated in a follow-up when the children were 8 years of age (108 boys and 77 girls). The children's perinatal dioxin exposure was estimated using the dioxin levels in their mothers' breast milk. The Colorado Learning Difficulties Questionnaire (CLDQ) was used to evaluate the children's learning difficulties. Math- and language-achievement scores were obtained using paper-based tests. Reading fluency was examined by having the children read passages in Vietnamese. RESULTS: In boys exposed to high levels of 2,3,7,8-tetrachlorodibenzodioxin (2,3,7,8-TetraCDD), CLDQ reading scores were significantly higher (worse), and language achievement scores were significantly lower. Boys exposed to high levels of 2,3,7,8-TetraCDD as well as high levels of the toxic equivalent (TEQ) of polychlorodibenzodioxins and polychlorodibenzofurans (PCDDs/Fs) had higher numbers of reading errors. Reading errors were higher and math achievement scores were lower with increasing concentrations of 1,2,3,4,7,8-HexaCDD and 1,2,3,4,6,7,8-HeptaCDD. In girls, no significant differences of any learning ability markers were found between high and low exposure groups to TEQ-PCDDs/Fs and these 3 congeners. CONCLUSIONS: Perinatal dioxin exposure may have adverse effects on the learning abilities of school children, especially boys.