Machine learning models for predicting unscheduled return visits of patients with abdominal pain at emergency department and validation during COVID-19 pandemic: A retrospective cohort study.

Machine learning (ML) models for predicting 72-hour unscheduled return visits (URVs) for patients with abdominal pain in the emergency department (ED) were developed in a previous study. This study refined the data to adjust previous prediction models and evaluated the model performance in future data validation during the COVID-19 era. We aimed to evaluate the practicality of the ML models and compare the URVs before and during the COVID-19 pandemic. We used electronic health records from Chang Gung Memorial Hospital from 2018 to 2019 as a training dataset, and various machine learning models, including logistic regression (LR), random forest (RF), extreme gradient boosting (XGB), and voting classifier (VC) were developed and subsequently used to validate against the 2020 to 2021 data. The models highlighted several determinants for 72-hour URVs, including patient age, prior ER visits, specific vital signs, and medical interventions. The LR, XGB, and VC models exhibited the same AUC of 0.71 in the testing set, whereas the VC model displayed a higher F1 score (0.21). The XGB model demonstrated the highest specificity (0.99) and precision (0.64) but the lowest sensitivity (0.01). Among these models, the VC model showed the most favorable, balanced, and comprehensive performance. Despite the promising results, the study illuminated challenges in predictive modeling, such as the unforeseen influences of global events, such as the COVID-19 pandemic. These findings not only highlight the significant potential of machine learning in augmenting emergency care but also underline the importance of iterative refinement in response to changing real-world conditions.

Machine learning models for predicting unscheduled return visits to an emergency department: a scoping review.

BACKGROUND: Unscheduled return visits (URVs) to emergency departments (EDs) are used to assess the quality of care in EDs. Machine learning (ML) models can incorporate a wide range of complex predictors to identify high-risk patients and reduce errors to save time and cost. However, the accuracy and practicality of such models are questionable. This review compares the predictive power of multiple ML models and examines the effects of multiple research factors on these models' performance in predicting URVs to EDs. METHODS: We conducted the present scoping review by searching eight databases for data from 2010 to 2023. The criteria focused on eligible articles that used ML to predict ED return visits. The primary outcome was the predictive performances of the ML models, and results were analyzed on the basis of intervals of return visits, patient population, and research scale. RESULTS: A total of 582 articles were identified through the database search, with 14 articles selected for detailed analysis. Logistic regression was the most widely used method; however, eXtreme Gradient Boosting generally exhibited superior performance. Variations in visit interval, target group, and research scale did not significantly affect the predictive power of the models. CONCLUSION: This is the first study to summarize the use of ML for predicting URVs in ED patients. The development of practical ML prediction models for ED URVs is feasible, but improving the accuracy of predicting ED URVs to beyond 0.75 remains a challenge. Including multiple data sources and dimensions is key for enabling ML models to achieve high accuracy; however, such inclusion could be challenging within a limited timeframe. The application of ML models for predicting ED URVs may improve patient safety and reduce medical costs by decreasing the frequency of URVs. Further research is necessary to explore the real-world efficacy of ML models.

In Situ Co-O Bond Reinforcement of the Artificial Cathode Electrolyte Interphase in Highly Delithiated LiCoO2 for High-Energy-Density Applications.

Highly delithiated LiCoO2 has high specific capacity (>200 mAh g-1); however, its degradation behavior causes it to have poor electrochemical performance and thermal instability. The degradation of highly delithiated LiCoO2 is mainly induced by oxygen vacancy migration and weakening of oxygen-related interactions, which result in pitting corrosion and fault formation on the surface. In this research, a coupling agent, namely, 3-aminopropyltriethoxysilane (APTES), was grafted onto the surface of LiCoO2 to form a cross-linking structure. Through the aza-Michael addition reaction, an oligomer formed from barbituric acid and bisphenol a diglycidyl ether diacrylate were reacted with the cross-linking APTES to form an artificial cathode electrolyte interphase (ACEI). The highly delithiated LiCoO2 containing the ACEI had considerably less degradation on the surface of the bulk material caused by oxygen release. The formation of the O1 phase was prevented in high delithiation and high-temperature operations. This research revealed that the ACEI reinforced the Co-O bond, which is crucial in preventing gas evolution and O1 phase formation. In addition, the ACEI prevents direct contact between the electrolyte and highly active surface of LiCoO2, thereby preventing the formation of a thick and high impedance traditional cathode electrolyte interphase. According to the present results, highly delithiated LiCoO2 containing the ACEI exhibited outstanding cycle retention and capacity at 55 °C as well as low heat capacity release in the fully delithiated state. The ACEI considerably protected and maintained the electrochemical performance of highly delithiated LiCoO2, which is suitable for high-energy-density applications, such as electric vehicles and power tools.

Effective Ru/CNT Cathode for Rechargeable Solid-State Li-CO2 Batteries.

An effective Ru/CNT electrocatalyst plays a crucial role in solid-state lithium-carbon dioxide batteries. In the present article, ruthenium metal decorated on a multi-walled carbon nanotubes (CNTs) is introduced as a cathode for the lithium-carbon dioxide batteries with Li1.5Al0.5Ge1.5(PO4)3 solid-state electrolyte. The Ru/CNT cathode exhibits a large surface area, maximum discharge capacity, excellent reversibility, and long cycle life with low overpotential. The electrocatalyst achieves improved electrocatalytic performance for the carbon dioxide reduction reaction and carbon dioxide evolution reaction, which are related to the available active sites. Using the Ru/CNT cathode, the solid-state lithium-carbon dioxide battery exhibits a maximum discharge capacity of 4541 mA h g-1 and 45 cycles of battery life with a small voltage gap of 1.24 V compared to the CNT cathode (maximum discharge capacity of 1828 mA h g-1, 25 cycles, and 1.64 V as voltage gap) at a current supply of 100 mA g-1 with a cutoff capacity of 500 mA h g-1. Solid-state lithium-carbon dioxide batteries have shown promising potential applications for future energy storage.

Controlling Ni2+ from the Surface to the Bulk by a New Cathode Electrolyte Interphase Formation on a Ni-Rich Layered Cathode in High-Safe and High-Energy-Density Lithium-Ion Batteries.

Ni-rich high-energy-density lithium ion batteries pose great risks to safety due to internal short circuits and overcharging; they also have poor performance because of cation mixing and disordering problems. For Ni-rich layered cathodes, these factors cause gas evolution, the formation of side products, and life cycle decay. In this study, a new cathode electrolyte interphase (CEI) for Ni2+ self-oxidation is developed. By using a branched oligomer electrode additive, the new CEI is formed and prevents the reduction of Ni3+ to Ni2+ on the surface of Ni-rich layered cathode; this maintains the layered structure and the cation mixing during cycling. In addition, this new CEI ensures the stability of Ni4+ that is formed at 100% state of charge in the crystal lattice at high temperature (660 K); this prevents the rock-salt formation and the over-reduction of Ni4+ to Ni2+. These findings are obtained using in situ X-ray absorption spectroscopy, operando X-ray diffraction, operando gas chromatography-mass spectroscopy, and X-ray photoelectron spectroscopy. Transmission electron microscopy reveals that the new CEI has an elliptical shape on the material surface, which is approximately 100 nm in length and 50 nm in width, and covers selected particle surfaces. After the new CEI was formed on the surface, the Ni2+ self-oxidation gradually affects from the surface to the bulk of the material. It found that the bond energy and bond length of the Ni-O are stabilized, which dramatically inhibit gas evolution. The new CEI is successfully applied in a Ni-rich layered compound, and the 18650- and the punch-type full cells are fabricated. The energy density of the designed cells is up to 300 Wh/kg. Internal short circuit and overcharging safety tests are passed when using the standard regulations of commercial evaluation. This new CEI technology is ready and planned for future applications in electric vehicle and energy storage.

Comparative Study of Li-CO2 and Na-CO2 Batteries with Ru@CNT as a Cathode Catalyst.

Alkali metal-carbon dioxide (Li/Na-CO2) batteries have generated widespread interest in the past few years owing to the attractive strategy of utilizing CO2 while still delivering high specific energy densities. Among these systems, Na-CO2 batteries are more cost effective than Li-CO2 batteries because the former uses cheaper and abundant Na. Herein, a Ru/carbon nanotube (CNT) as a cathode material was used to compare the mechanisms, stabilities, overpotentials, and energy densities of Li-CO2 and Na-CO2 batteries. The potential of Na-CO2 batteries as a viable energy storage technology was demonstrated.

A Machine Learning Model for Predicting Unscheduled 72 h Return Visits to the Emergency Department by Patients with Abdominal Pain.

Seventy-two-hour unscheduled return visits (URVs) by emergency department patients are a key clinical index for evaluating the quality of care in emergency departments (EDs). This study aimed to develop a machine learning model to predict 72 h URVs for ED patients with abdominal pain. Electronic health records data were collected from the Chang Gung Research Database (CGRD) for 25,151 ED visits by patients with abdominal pain and a total of 617 features were used for analysis. We used supervised machine learning models, namely logistic regression (LR), support vector machine (SVM), random forest (RF), extreme gradient boosting (XGB), and voting classifier (VC), to predict URVs. The VC model achieved more favorable overall performance than other models (AUROC: 0.74; 95% confidence interval (CI), 0.69-0.76; sensitivity, 0.39; specificity, 0.89; F1 score, 0.25). The reduced VC model achieved comparable performance (AUROC: 0.72; 95% CI, 0.69-0.74) to the full models using all clinical features. The VC model exhibited the most favorable performance in predicting 72 h URVs for patients with abdominal pain, both for all-features and reduced-features models. Application of the VC model in the clinical setting after validation may help physicians to make accurate decisions and decrease URVs.

Interface Interaction Behavior of Self-Terminated Oligomer Electrode Additives for a Ni-Rich Layer Cathode in Lithium-Ion Batteries: Voltage and Temperature Effects.

Self-terminated oligomer additives synthesized from bismaleimide and barbituric acid derivatives improve the safety and performance of lithium-ion batteries (LIBs). This study investigates the interface interaction of these additives and the cathode material. Two additives were synthesized by Michael addition (additive A) and aza-Michael addition (additive B). The electrochemical performances of bare and modified LiNi0.6Mn0.2Co0.2O2 (NMC622) materials are studied. The cycling stability and rate capability of NMC622 considerably improve on surface modification with additive B. According to the differential scanning calorimetry results, the exothermic heat of fully deliathiated NMC622 is dramatically decreased through surface modification with both additives. The electrode surface kinetics and interface interaction phenomena of the additives are determined through surface plasma resonance measurements in operando gas chromatography-mass spectroscopy (GCMS) and in situ soft X-ray absorption spectroscopy (XAS). The binding rate constant of additive B onto NMC622 particles is 1.2-2.3 × 104 M-1 s-1 in the temperature range of 299-311 K, which is ascribed to the strong binding affinity toward the electrode surface. This affinity enhances Li+ diffusion, which allows the electrode modified by additive B to provide high electrochemical performance with superior thermal stability. In operando GCMS reveals that gas evolution due to the electrolyte degradation at the NMC622 surface contributes to safety hazards in the bare NMC622 material. In situ soft XAS indicates the occurrence of structural transformation in the bare NMC622 material after it is fully charged and at elevated temperatures. The NMC622 material is stabilized by incorporating additives. The unique performance of additive B can be attributed to its linear structure that allows superior electrode surface adhesion compared with that of additive A. Therefore, this study presents an optimized working principle of self-terminated oligomers, which can be developed and applied to improve the safety and performance of LIBs.