Large Datasets Research Articles

Revolutionizing Cardio Vascular Disease Diagnosis Through Lasso Regression Model

Cardiovascular diseases (CVDs) are among the most prevalent and fatal health conditions globally, necessitating early and accurate diagnosis to mitigate risk and enhance treatment outcomes. Traditional diagnostic methods often rely on a combination of clinical assessments and static risk models, which can be constrained by their inability to handle the complexity and interrelationships of multiple risk factors. In this research, we propose a novel approach to revolutionizing cardiovascular disease diagnosis through the application of the Lasso (Least Absolute Shrinkage and Selection Operator) regression model, a powerful machine learning technique for feature selection and predictive modeling. The Lasso regression model is particularly advantageous for high-dimensional medical datasets, where multiple clinical features may be correlated. By applying Lasso, we aim to address two critical challenges in cardiovascular disease prediction: identifying the most relevant predictors from large and complex datasets and reducing model over-fitting, which is common when working with a large number of co-variates. Our results demonstrated that the Lasso model significantly improves prediction accuracy when compared to traditional logistic regression and other machine learning models. Lasso regression has the potential to revolutionize cardiovascular disease diagnosis by providing a data-driven, efficient, and interpretable solution that bridges complex medical datasets with practical clinical decision-making. Keywords: Cardio vascular diseases (CVD), Machine learning models, Losso regression model.

SOCIAL MEDIA FAKE ACCOUNT IDENTIFICATION USING ML

The proliferation of social media platforms has led to an increase in the creation of fake accounts. These accounts are used for various malicious activities, such as spreading false information, phishing, and identity theft. As a result, there is a growing need for effective methods to identify and eliminate fake accounts. This paper proposes a machine learning-based approach for social media fake account identification. This paper proposes a machine learning-based approach to identify fake accounts on social media platforms. Our method leverages a combination of feature extraction techniques and supervised learning algorithms to classify accounts as genuine or fake. We collect a large dataset of labeled social media accounts, including both genuine and fake profiles, and extract features from various sources such as profile information, network behavior, and content analysis. We experiment with multiple machine learning models, including Support vector machines(SVM),K-Nearest Neighbors Algorithm(KNN),Random forest, Logistic Regression & Artificial Neural Network(ANN) to evaluate their performance in identifying fake accounts. Our proposed method has significant implications for social media platform operators, policymakers, and researchers seeking to combat fake accounts and maintain online trust. The approach can be integrated with existing social media moderation tools to enhance the accuracy and efficiency of fake account identification. Key Words: Support vector machines(SVM),K-Nearest Neighbors Algorithm(KNN),Random forest, Logistic Regression & Artificial Neural Network(ANN),Python.

SENTIMENT ANALYSIS OF PUBLIC OPINION ON PRESIDENTIAL ADVISORY APPOINTMENTS USING NAIVE BAYES CLASSIFICATION

Social media platforms such as Twitter, Facebook, and YouTube have become significant channels for public discourse, where users freely express opinions, including negative sentiments and hate speech. To better understand public opinion, particularly in politically charged contexts, sentiment analysis can classify user comments as either positive or negative. This study aims to analyze public sentiment regarding the formation of a special advisory team for President Jokowi, using a sentiment classification approach. The study employed a Naïve Bayes classifier to analyze sentiment from 3,000 comments gathered from Twitter, Facebook, and YouTube. The dataset was divided into 80% training data (used to train the model with known sentiment) and 20% test data. The Naïve Bayes algorithm was chosen for its simplicity and effectiveness in handling large datasets in text classification tasks. The Naive Bayes classification on sentiment analysis of public opinion regarding the appointment of presidential advisors achieved an overall accuracy of 71% in classifying the test data. Negative sentiment was classified with an accuracy of 71%, while positive sentiment was classified with an accuracy of 70%. The results demonstrate that the Naïve Bayes classifier is a viable method for sentiment analysis in political discourse, although the model's performance indicates room for improvement. The novelty of this research lies in its focus on sentiment analysis of public opinion specifically related to presidential advisory appointments, an area not yet extensively explored in sentiment analysis studies. This study contributes to the field by providing insights into the public’s perception of political decisions using machine learning techniques. The implications for future research include refining classification methods for better accuracy and applying the model to other political or governmental topics.

Parameter Dependency of Electrochemical Reduction of CO2 in Acetonitrile - A Data Driven Approach.

The electrochemical CO2 reduction reaction (CO2RR) is a promising technology for the utilization of captured CO2. Though systems using aqueous electrolytes is the state-of-the-art, CO2RR in aprotic solvents are a promising alternative that can avoid the parallel hydrogen evolution reaction (HER). While system parameters, such as electrolyte composition, electrode material, and applied potential are known to influence the reaction mechanism, there is a lack of intuitive understanding as to how. We show that by using multivariate data analysis on a large dataset collected from the literature, namely random forest modelling, the most important system parameters can be isolated for each possible product. We find that water content, current density, and applied potential are powerful determinants in the reaction pathway, and therefore in the Faradaic efficiency of CO2RR products.

Addressing the Global Logistics Performance Index Rankings with Methodological Insights and an Innovative Decision Support Framework

This study examines the Logistics Performance Index (LPI) rankings developed by the World Bank from a methodological perspective and proposes an alternative decision support framework. LPI serves as an interactive tool that helps countries identify challenges, innovative solutions, and opportunities in their trade and logistics sectors. In this study, the efficiency of logistics operations in 118 countries was evaluated using an integrated multi-criteria decision-making (MCDM) model objectively weighted by the Entropy method. Countries were ranked using the MCRAT, SAW, TOPSIS, and FUCA methods. According to the findings, large datasets provide more robust insights for sensitivity analyses, and wider weighting coefficient combinations make the data more meaningful. In addition, it is suggested to use low-compensation methods instead of classical additive methods for LPI. Unlike other studies in literature, this research applied an innovative sensitivity analysis to test the robustness of the model and comprehensively examined the effects of weighting techniques based on over 2500 different MCDM results. The findings suggest that the FUCA method should be recommended to decision-makers for calculating LPI rankings due to its simplicity, practicality, low compensatory power, and low sensitivity. This study offers methodological improvements when evaluating logistics performance and provides significant contributions to decision-making processes. The findings are expected to provide a valuable resource for policymakers and businesses in understanding a country’s position in global competition, as well as serving as a reference for researchers evaluating the logistics performance of countries.

Travel route recommendation with a trajectory learning model

This study addresses a critical issue in location-based services: travel route recommendation. It leverages historical trajectory data to predict the actual route on a road network from a starting point to a destination, given a specific departure time. However, capturing the latent patterns in complex trajectory data for accurate route planning presents a significant challenge. Existing route recommendation methods commonly face two major problems: first, inadequate integration of multi-source data, which fails to fully consider the potential factors affecting route choice; and second, limited capability to capture road network characteristics, which restricts the effective application of node features and negatively impacts recommendation accuracy. To address these issues, this research introduces a Trajectory Learning Model for Route Recommendation (TLMR) based on deep learning techniques. TLMR enhances the understanding of user route choice behavior in complex environments by integrating multi-source data. Moreover, by incorporating road network features, TLMR more effectively captures and utilizes the structural and dynamic information of the road network. Specifically, TLMR first employs a Position-aware Graph Neural Network to learn features of intersections from the road network, incorporating context features like weather and traffic conditions. Then, it integrates this information through neural networks to predict the next intersection. Finally, a beam search algorithm is applied to generate and recommend multiple candidate routes. Extensive experiments on four large real-world datasets demonstrate that TLMR outperforms existing methods in four key performance metrics. These results prove the effectiveness and superiority of TLMR in route recommendation.

DeepRFQC: automating quality control for P-wave receiver function analysis using a U-net inspired network

This paper introduces DeepRFQC, an automated method for quality control in P-wave receiver function analysis. Leveraging a U-Net inspired deep learning model, which has previously shown promise in denoising and phase detection, DeepRFQC efficiently distinguishes usable from noisy receiver functions. We examine a Proterozoic Trans-Hudson Orogen dataset from northern Canada, including seismic events from 1990 to 2023, which is expanded for training purposes by data augmentation techniques. With 1,508,449 trainable parameters, the DeepRFQC model attains a commendable 96.6% validation accuracy, on a test dataset from the X5 seismic network; tests on stations from different tectonic environments indicate that the model is effective even in environments very different from the training set. Validation through the H-κ stacking method shows consistent and plausible results. As manual quality control is a major bottleneck in receiver-function processing, automated methods such as this one will allow for efficient examination of large data sets.

Conceptual advances in petrophysical inversion techniques: The synergy of machine learning and traditional inversion models

Petrophysical inversion techniques are essential for estimating reservoir properties such as porosity, permeability, and fluid saturation, which are critical for optimizing hydrocarbon recovery and reservoir management. Traditionally, deterministic and stochastic inversion methods have relied on well log data, core samples, and seismic information to generate models of subsurface properties. However, these techniques face limitations in handling large datasets, nonlinear relationships, and uncertainties in complex reservoir environments. Recent advances in machine learning (ML) offer an opportunity to enhance these traditional methods by providing a data-driven approach capable of learning patterns and refining inversion models. This review explores the potential synergy between traditional petrophysical inversion techniques and machine learning algorithms to create a new paradigm for reservoir parameter estimation and analysis. Machine learning’s ability to process large volumes of diverse data and capture nonlinear relationships complements the physics-based constraints of traditional inversion methods. By combining these two approaches, hybrid models can provide improved accuracy in predicting reservoir properties and handling data gaps or uncertainties. The review discusses specific case studies where machine learning-assisted inversion models have been successfully applied, such as the estimation of porosity and permeability, seismic inversion for lithology identification, and fluid saturation modeling. These hybrid models demonstrate enhanced predictive accuracy, faster computational efficiency, and real-time adaptability, transforming reservoir characterization and management. The integration of machine learning with traditional inversion techniques represents a major advancement in petrophysical analysis. This synergy holds the potential to revolutionize reservoir parameter estimation, making it more accurate, robust, and responsive to real-time data, ultimately improving decision-making in reservoir engineering and maximizing hydrocarbon recovery. Keywords: Petrophysical Inversion, Machine Learning, Traditional Inversion, Models.

Effect of L-Carnitine Level on the Risk of Neuromyelitis Optica Spectrum Disorders: A Two-Sample Mendelian Randomization Study.

Previous research has often focused on studying the CNS damage in neuromyelitis optica spectrum disorders (NMOSD), while the role of the peripheral blood in the development of NMOSD is also of significant importance. The relationship between metabolites in blood and cerebrospinal fluid (CSF) with neuroimmune is receiving increasing attention. L-carnitine, whose astrocytic accumulation is associated with neuroinflammation and demyelination, may participate in the pathogenesis of NMOSD. However, whether circulating L-carnitine level has a causal effect on NMOSD risk needs elucidation. With large data sets now available, we used two-sample Mendelian randomization (MR) to determine whether circulating L-carnitine level is causally associated with the risk of NMOSD. Genetic variants associated with circulating L-carnitine levels were derived from a genome-wide association study (GWAS) of 7797 individuals from TwinsUK and KORA F4 cohorts. NMOSD summary statistics, including 215 cases and 1244 controls, were obtained from a separate GWAS. Subgroup analyses included aquaporin-4 (AQP4)-IgG-seropositive NMOSD (132 cases) and AQP4-IgG-seronegative NMOSD (83 cases). We used two-sample MR to explore associations between circulating L-carnitine levels and NMOSD risk, as well as its seropositive and seronegative subtypes. 16 SNPs (single nucleotide polymorphisms) were significantly associated with circulating L-carnitine level (P < 5 × 10-8), all of which were independent and available in the NMOSD dataset, after 1 SNP removed for being palindromic with intermediate allele frequencies in harmonization. Finally, a high circulating L-carnitine level conferred a protective effect against combined NMOSD (OR = 2.216 × 10-4, 95% confidence interval [CI] = 2.335 × 10-7-2.104 × 10-1, P = 0.0161) as well as AQP4-IgG-seronegative NMOSD (OR = 7.678 × 10-7, 95% CI = 2.233 × 10-11-2.640 × 10-2, P = 0.0082). There is no causal effect on AQP4-IgG-seropositive NMOSD risk (OR = 5.471 × 10-3, CI = 1.090 × 10-6-27.465, P = 0.2798) by circulating L-carnitine. Results remained positive and robust after the horizontal pleiotropy test, heterogeneity test, and Bonferroni test. In the reverse MR analysis, there was no causal effect of NMOSD and its subtypes on circulating L-carnitine levels. Our findings suggest that higher circulating L-carnitine levels may reduce the risk of NMOSD, particularly in AQP4-IgG-seronegative patients. L-carnitine could serve as a valuable biomarker and potential therapeutic target for NMOSD, especially in cases without AQP4-IgG. The genetic evidence from this study supports further exploration of L-carnitine's role in managing NMOSD.

Dynamic pricing and demand forecasting: Integrating time-series analysis, regression models, machine learning, and competitive analysis

Abstract. Dynamic pricing is a critical strategy for businesses seeking to optimize revenue and stay competitive in fluctuating markets. This paper explores the integration of various demand forecasting techniques, including time-series analysis, regression models, and machine learning algorithms, with competitive analysis methodologies to enhance dynamic pricing strategies. Time-series analysis focuses on decomposing data into trend, seasonality, and random fluctuations, using ARIMA models for accurate demand prediction. Regression models delve into the complexities of variable interactions, extending beyond linear relationships to include advanced techniques like Ridge, Lasso, and Elastic Net regression. Machine learning algorithms, such as decision trees, random forests, gradient boosting, and neural networks, revolutionize demand forecasting by uncovering complex patterns in large datasets. Competitive analysis incorporates market scanning, price elasticity estimation, and competitor behavior modeling to inform dynamic pricing decisions. Optimization algorithms, including linear programming, genetic algorithms, and simulated annealing, are employed to refine pricing strategies, while revenue management techniques like yield management and overbooking ensure maximum revenue from perishable goods and services. This comprehensive approach enables businesses to dynamically adjust prices in real-time, maintaining competitiveness and maximizing profitability in an ever-evolving market landscape.

Implementation of Distributed Machine Learning in Cloud Calculation: Evidence from Consumer Behavior Prediction

Abstract. In the field of machine learning, distributed machine learning (DML) has gained massive popularity in research and appeared in a wide range of applications including transportation, medicine, and business. Distributed machine learning showed huge potential in processing large amounts of data while keeping the data private. In this paper, a distributed implementation of a deep neural network (DNN) with a large dataset from a telecom company, aiming to predict consumer behavior based on usage data is carried out. Various datasets with distinct heterogeneity settings were applied to the network, simulating a real-world cloud application scenario. Comparisons between both the centralized model and the distributed models were made to analyze the impact on model performance with device heterogeneity and data heterogeneity. These result indicate that under certain circumstances, DML models showed better accuracy and lower loss compared to the centralized model. The paper provided some insights and made some assumptions on how the data and the parameters affect the model performance.

GUing: A Mobile GUI Search Engine using a Vision-Language Model

Graphical User Interfaces (GUIs) are central to app development projects. App developers may use the GUIs of other apps as a means of requirements refinement and rapid prototyping or as a source of inspiration for designing and improving their own apps. Recent research has thus suggested retrieving relevant GUI designs that match a certain text query from screenshot datasets acquired through crowdsourced or automated exploration of GUIs. However, such text-to-GUI retrieval approaches only leverage the textual information of the GUI elements, neglecting visual information such as icons or background images. In addition, retrieved screenshots are not steered by app developers and lack app features that require particular input data. To overcome these limitations, this paper proposes GUing, a GUI search engine based on a vision-language model called GUIClip, which we trained specifically for the problem of designing app GUIs. For this, we first collected from Google Play app introduction images which display the most representative screenshots and are often captioned (i.e. labelled) by app vendors. Then, we developed an automated pipeline to classify, crop, and extract the captions from these images. This resulted in a large dataset which we share with this paper: including 303k app screenshots, out of which 135k have captions. We used this dataset to train a novel vision-language model, which is, to the best of our knowledge, the first of its kind for GUI retrieval. We evaluated our approach on various datasets from related work and in a manual experiment. The results demonstrate that our model outperforms previous approaches in text-to-GUI retrieval achieving a Recall@10 of up to 0.69 and a HIT@10 of 0.91. We also explored the performance of GUIClip for other GUI tasks including GUI classification and sketch-to-GUI retrieval with encouraging results.

Comparative Analysis of Distributed and Parallel Machine Learning: Principles, Models, and Applications

Abstract. Contemporarily, the explosive growth of technology and the growth of the complexity of machine learning models facilitated the advancement of parallel and distributed machine learning. This study compared and analyzed the two paradigms from three perspectives: principles, models, and applications. This research analyzed the basic principle and some important models of parallel and distributed machine learning, mainly focusing on their operating mechanism, scalability, and performance metrics. The analysis includes how these methods can help to deal with large datasets and complex models. According to the evaluations, distributed machine learning is better in managing massive data and achieving model convergence, while parallel machine learning performs well in processing complicated models quickly. These results provided opinions on each methods advantages and disadvantages and suggestions on how to choose between the two methods in different situations. This research also stressed the importance of understanding these differences in advancing machine learning and promoting innovation in data-intensive fields.

Comparative Analysis of Machine Learning, Decision Trees, and K-Nearest Neighbors for Heart Disease Prediction

Abstract. This study investigates the efficacy of Machine Learning (ML), Decision Trees, and K-Nearest Neighbors (KNN) techniques in predicting heart disease, aiming to identify their strengths and limitations. ML models are effective in detecting complex patterns and delivering evolving predictions from large datasets but require high-quality data and can be challenging to interpret. Decision Trees provide clear, understandable decision rules, which is beneficial in clinical settings, yet they are susceptible to overfitting and instability. KNN, valued for its simplicity and flexibility, classifies heart disease based on similarity but struggles with high computational costs and sensitivity to noisy data. Experimental results indicate that each model has distinct advantages: ML excels in pattern recognition, Decision Trees offer interpretability, and KNN handles diverse data effectively. However, each also faces challenges that impact performance, such as data quality issues for ML, overfitting for Decision Trees, and computational demands for KNN. The study suggests that balancing these strengths and weaknesses is crucial for optimizing heart disease prediction models. Future research should explore hybrid approaches that combine these models' advantages while addressing their respective limitations to improve predictive accuracy and practical application in real-world scenarios.

THE APPLICATION OF DATA MINING AND OLAP TECHNOLOGIES IN HEALTHCARE

The article examines the application of Data Mining and real-time Online Analytical Processing (OLAP) in healthcare. The calculation of analytical indicators and solving planning problems in healthcare are complex and multi-stage processes. The extraction of knowledge from large volumes of data accumulated in databases is referred to as Data Mining. Today, databases are measured in terabytes, containing strategically significant information. One of the most critical issues is determining methods for discovering important information within such vast amounts of data. Data Mining technology, by utilizing a set of analytical tools, helps to uncover relationships within large datasets and make predictions about future trends. Simply defined, Data Mining automatically identifies relevant patterns within databases. For many years, analysts manually recorded statistically significant relationships within databases. Today, Data Mining automates this process. The goal of Data Mining is to create decision-making models that predict future behavior based on the analysis of past actions. With the advent of computers and the rise of the digital information age, there have been changes in both the scale and complexity of existing problems in terms of analytical analysis. Advances in the storage, management, and examination of data have led to the emergence of a new field that promotes the development of bioinformatics and introduces analytical analysis and computational problems in the fields of biology and medicine. Keywords: Data Mining, OLAP, database, factor analysis, healthcare network, multidimensional models.

Comparison Distributed and Parallel Machine Learning: Evidence from Models, Principles and Application Scenarios

Abstract. Contemporarily, the demand for processing large-scale data has been rapidly increasing, prompting continuous advancements in the field of machine learning. This study examines the state of parallel and distributed machine learning at present, both of which aim to enhance computational efficiency when dealing with large datasets and demonstrate promising applications across various domains. As the scale of data escalates, traditional single-machine learning methods are becoming increasingly inadequate, which lead to the emergence of parallel and distributed machine learning. These approaches enable substantial computations to be performed more efficiently through the collaboration of multi-core CPUs or multiple computing nodes. This research conducts an in-depth analysis of the inherent challenges associated with these methods, including data transmission latency, synchronization requirements, and user privacy concerns. Ultimately, this research emphasizes the tremendous potential of both methods for future applications, a potential that is bolstered by ongoing advancements in hardware and algorithm optimization. These results provide valuable insights for practitioners in the field and offers guidance for future research directions.

Robust Biometric Verification Using Phonocardiogram Fingerprinting and a Multilayer-Perceptron-Based Classifier

Recently, a new set of biometric traits, called medical biometrics, have been explored for human identity verification. This study introduces a novel framework for recognizing human identity through heart sound signals, commonly referred to as phonocardiograms (PCGs). The framework is built on extracting and suitably processing Mel-Frequency Cepstral Coefficients (MFCCs) from PCGs and on a classifier based on a Multilayer Perceptron (MLP) network. A large dataset containing heart sounds acquired from 206 people has been used to perform the experiments. The classifier was tuned to obtain the same false positive and false negative misclassification rates (equal error rate: EER = FPR = FNR) on chunks of audio lasting 2 s. This target has been reached, splitting the dataset into 70% and 30% training and testing non-overlapped subsets, respectively. A recurrence filter has been applied to also improve the performance of the system in the presence of noisy recordings. After the application of the filter on chunks of audio signal lasting from 2 to 22 s, the performance of the system has been evaluated in terms of recall, specificity, precision, negative predictive value, accuracy, and F1-score. All the performance metrics are higher than 97.86% with the recurrence filter applied on a window lasting 22 s and in different noise conditions.

The DOSE index in chronic obstructive pulmonary disease: evaluating healthcare costs

Background and objectivesThe DOSE index, which incorporates Dyspnea, Obstruction, Smoking, and Exacerbations, is a widely used tool for assessing the severity and prognosis of Chronic Obstructive Pulmonary Disease (COPD). In addition to risk assessment, it has potential clinical utility in predicting healthcare costs, which are primarily driven by exacerbations. While several indices, such as the BODE (Body-mass index, Obstruction, Dyspnea, Exercise) and ADO (Age, Dyspnea, Obstruction) indices, exist for risk prediction, there is a lack of dedicated tools for forecasting healthcare costs. This study explores the potential of the DOSE index compared to other indices, including BODE, ADO, and the Charlson Comorbidity Index (CCI), for this purpose.Materials and methodsThis cross-sectional retrospective study analyzed data from 396 COPD cases. We examined associations between the DOSE index, BODE index, ADO index, CCI, and healthcare costs, including hospitalizations and emergency room treatments. Healthcare costs were categorized as direct medical expenses.ResultsSignificant associations were observed between the DOSE index and various healthcare parameters. DOSE quartiles showed strong correlations with outpatient visits (p = 0.013) and outpatient medical expenses (p = 0.011). In addition, hospitalization frequency, duration, and associated costs were significantly correlated with higher DOSE quartiles (p < 0.001). A significant difference was found when comparing DOSE quartiles between patients with high (CCI ≥ 3) and low (CCI < 3) comorbidity scores (p = 0.018). The DOSE index outperformed other indices, likely due to its inclusion of exacerbations, a key driver of healthcare costs.ConclusionThe DOSE index demonstrates potential in predicting healthcare costs, particularly due to its inclusion of exacerbation frequency. This study highlights the importance of considering exacerbations alongside traditional risk factors for more accurate cost forecasting in COPD management. Our findings suggest that the DOSE index may be a valuable tool in both clinical and economic assessments of COPD patients, though further research is warranted to validate these findings in larger datasets.

Predictors and Predictive Score of In-Hospital Mortality in Diabetic Ketoacidosis: A Retrospective Cohort Study

Background and Objectives: Diabetic ketoacidosis (DKA) is a critical complication of diabetes mellitus (DM). The primary objective of this study was to identify relevant clinical and biochemical predictors and create a predictive score for in-hospital DKA mortality. Materials and Methods: A 6-year retrospective cohort study of adult patients diagnosed with DKA and admitted to Chiang Mai University Hospital, a tertiary care center in Chiang Mai, Thailand, from 1 January 2015 to 31 December 2021, was conducted. Baseline clinical data and laboratory investigations were collected. The primary outcome was in-hospital mortality. Multivariable logistic regression analysis, clustered by type of diabetes, was performed to identify significant predictors. A predictive risk score was created using significant predictive factors identified by multivariable analysis. The results were presented as odds ratios (ORs) and 95% confidence intervals (CIs), with a significant p-value set at &lt;0.05. Results: Ninety-three patients diagnosed with DKA were included in the study. Ten patients died during admission. Significant predictors for in-hospital mortality of DKA included age &gt; 55 years (OR 7.8, p = 0.007), female gender (OR 3.5, p &lt; 0.001), anion gap &gt; 30 mEq/L (OR 2.6, p = 0.003), hemoglobin levels &lt; 10 g/dL (OR 16.9, p &lt; 0.001), and the presence of cardiovascular disease (OR 1.3, p = 0.046). The predictive risk score ranged from 1 to 14 for low risk, and 14.5–23.5 for high risk of in-hospital mortality. The predictive performance of the scoring system was 0.82 based on the area under the curve, with a sensitivity of 73.8% and specificity of 96.4%. Conclusions: Multiple clinical and biochemical factors, along with a predictive risk score, could assist in predicting in-hospital mortality of DKA and serve as a guide for physicians to identify patients at high risk. Nevertheless, as the predictive score was internally validated with data from a single institution, external validation in diverse healthcare settings with larger datasets or prospective cohorts is crucial to confirm the model’s generalizability and predictive accuracy.

Exploiting Friedel pairs to interpret scanning 3DXRD data from complex geological materials

The present study introduces a processing strategy for synchrotron scanning 3D X-ray diffraction (s3DXRD) data, aimed at addressing the challenges posed by large, highly deformed, polyphase materials such as crystalline rocks. Leveraging symmetric Bragg reflections known as Friedel pairs, our method enables diffraction events to be precisely located within the sample volume. This method allows for fitting the phase, crystal structure and unit-cell parameters at the intra-grain scale on a voxel grid. The processing workflow incorporates several new modules, designed to (i) efficiently match Friedel pairs in large s3DXRD datasets containing up to 108 diffraction peaks; (ii) assign phases to each pixel or voxel, resolving potential ambiguities arising from overlap in scattering angles between different crystallographic phases; and (iii) fit the crystal orientation and unit cell locally on a point-by-point basis. We demonstrate the effectiveness of our technique on fractured granite samples, highlighting the ability of the method to characterize complex geological materials and show their internal structure and mineral composition. Additionally, we include the characterization of a metal gasket made of a commercial aluminium alloy, which surrounded the granite sample during experiments. The results show the effectiveness of the technique in recovering information about the internal texture and residual strain of materials that have undergone high levels of plastic deformation.