Complete Data Research Articles

Distribution of Ocular Biometry and Intraocular Lens Power in Patients Undergoing Cataract Surgery in an Eye Camp in Northern Pakistan

Purpose: To report ocular biometric data of patients undergoing cataract surgery in an eye camp in a city of Northern area of Pakistan. Design: Cross-sectional descriptive study. Place and Duration of Study: A surgical eye camp was set up in the month of September 2022 for 6 days in a private hospital in Skardu. Methods: A total of 119 patients were selected with a non-probability consecutive sampling technique. Ocular biometry of permanent local residents with significant cataractswas included. Incomplete biometry data and coexisting diseases like corneal opacity, pterygium, and severe dry eye causing errors in biometry were excluded. Keratometry was done with an automated refractor-keratometer. Axial length was calculated with an A-scan (ultrasound contact method) and intraocular lens (IOL) power was calculated. Results: A total of 119 patients were selected for biometry among 3908 individuals visiting an eye camp. The mean axial length was 22.13 ± 1.25 SD mm. The mean keratometry readings (K1) was 43.28 ± 1.83 Diopter and (K2) 44.26 ± 1.67 D. The mean intraocular lens (IOL) power was 24.78 ± 3.43 D. The mean K readings were more in males as compared to females (p-value <0.005). There was no significant difference in axial length and IOL power among genders. Conclusion: This study gives normative data on ocular biometric parameters of patients undergoing cataract surgery at high altitudes in Skardu, Pakistan. These biometric values can be used as reference values in the local population

Impact of supportive oncology (SO) visits on healthcare utilization.

199 Background: Palliative care (PC) has demonstrated a reduction in acute care utilization for health systems. PC input for patients with cancer has been associated with fewer emergency room visits, hospitalizations and hospital deaths in the last 30 days of life. Less is known about the upstream effects of PC referral on healthcare utilization (HCU). We examined HCU among patients in the 90 days after first SO appointment and predictors of high HCU. Methods: Patients ≥18 years with first SO consult between 01/01/2021-06/30/2023 were included. Patients who died within 90 days of their first SO appointment and incomplete utilization data were censored. Data on patient demographics, acute healthcare and hospice utilization, cancer treatment and mortality was extracted from the electronic health record (EHR). Number of ED visits in 90 days before and after first SO visit were recorded. We used descriptive statistics to describe patients’ sociodemographics, clinical characteristics and healthcare utilization. High HCU was defined as patients in the top 10% of ED visits and IP stays. Logistic regression was employed to assess whether being a high utilizer is associated with baseline characteristics. Results: We included 859 patients. On average, there were approximately 0.12, 95%CI: [0.07,0.17] more ED visits in the 90 days before the first SO appointment compared to 90 days after. There was no difference in hospitalization rates before and after the first SO appointment. Overall, 45 patients met the criteria for high HCU. The median number of ED visits and hospitalizations of high utilizers was 6 visits, compared to 1 visit for remainder of the population. There was a significant association with ethnicity with those identified as Hispanic (60%) being high HCU. There was no significant association between age, insurance status, cancer type, primary language and high HCU. High utilizers were 189% more likely to die in hospital although this did not meet statistical significance (p= 0.242). Conclusions: The findings reveal significant variations in healthcare utilization, particularly among Hispanic patients, indicating a need for targeted interventions to address these disparities. Despite no observed changes in hospitalization rates, the study emphasizes the need for further research to understand the underlying factors contributing to high healthcare utilization. Future efforts should focus on developing tailored strategies to optimize PC referrals and enhance patient outcomes across diverse populations. Race/ethnicity by utilization group. Category <90th Percentile Utilization (n=814) % >90th Percentile Utilization (n=45) % P value Race (%) <0.001 Asian 97 (11) 0 (0) Black or African American 173 (21) 8 (18) Hispanic 198 (24) 27 (60) Other 50 (6) 3 (7) White 296 (36) 7 (16)

A smart agriculture information system delivering research data for adoption by the Australian grains industry

Online Farm Trials (OFT) is a bespoke online information system providing current and historical grains trials research data and information in Australia. It represents a technology that supports smart agriculture by enabling discoveries in the data. Its impact on users has been established, supporting industry knowledge and decision making and leading to improved work practices. The aim of this research is to identify the contribution of OFT, as a tool that supports smart agriculture, and to explore current adoption, barriers and the opportunities for expansion of the system in the Australian grains industry, as perceived by industry experts. In-depth, semi-structured interviews were conducted with leaders and innovators from the agriculture industry (N = 16). Thematic analysis of the qualitative data reveals widespread value of OFT as an important resource for accessing industry-dedicated, current and historical research trials data. Barriers to adoption were identified which included incomplete data, and a lack of industry-wide awareness. Opportunities to consolidate and expand the data, together with improved industry awareness, are essential to improving industry-wide adoption. Key learnings were identified in the research that has benefit for the industry, and consideration for the adoption of technology in smart agriculture. This includes functionality, audience, completeness of information/data, consideration of new technologies and the promotion of the system supported by real world examples to encourage industry adoption. Recommendations to support the design and development of digital platforms in agriculture, in the future, are also offered.

Automated Sample Annotation for Diabetes Mellitus in Healthcare Integrated Biobanking

Healthcare integrated biobanking describes the annotation and collection of residual samples from hospitalized patients for research purposes. The central idea of the current work is to establish an automated workflow for sample annotation, selection and storage for diabetes mellitus.This is challenging due to incomplete data at the time of sample selection. The study evaluates a machine learning (ML) and natural language processing (NLP) based two-step procedure for timely and precise sample annotation for diabetes mellitus.Electronic health record data of 785 persons were extracted from the hospital information system. In the first step, a conditional inference forest (CIF) model was trained and tested based on laboratory values from the first 72hours of the hospital stay using test- (n=550) and training data sets (n=235). Performance was compared with a simple laboratory cut-off classifier (LCC) and a logistic regression (LR) model. Algorithms based on laboratory values, ICD-10 codes or information from discharge summaries extracted by a natural language processing software (NLP-DS) were evaluated as a second (review) step designed to increase the precision of annotations.For the first step, recall/precision/F1-score/accuracy were 71%/86%/0.78/0.82 for CIF and 77%/70%/0.74/0.75 for LR compared to 73%/68%/0.70/0.72 for LCC. NLP-DS was the best performing second (review) step (93%/100%/0.97/0.97). Combining first-step models with NLP-DS increased precision to 100% for all procedures (66%/100%/0.80/0.85 for CIF&NLP-DS 72%/100%/0.84/87.2 for LR&NLP-DS and 66%/100%/0.80/0.85 for LCC&NLP-DS). The number of samples removed by the NLP-DS analysis was higher for LR&NLP-DS and LCC&NLP-DS(removal rate 35% and 38% of initially selected samples) compared to CIF&NLP-DS (removal rate of 20%).The developed two-step procedure is an efficient implementable method for timely and precise annotation of samples from diabetic hospitalized patients.

Learning and Development Program’s Admission Requirements and Student Engagement as Success Determinants in Accountancy

This study focuses on the admission criteria of a bachelor of accountancy (BAcc) degree program as one of many factors influencing students’ success in graduating in the regulated time frames. The present study used constructivist learning and student engagement theories as connected lenses to explore the effect of admission criteria on first-year-level students’ academic performance. The research findings emanate from a mixed methods methodology using logistic regression, one of the choice models in the International Business Machines (IBM) Statistical Package for Social Scientists (SPSS), and semi-structured interviews. A sample of 43 was randomly selected from the list of 91 students with NSC-level accounting to match the number of 43 students who did not do NSC-level accounting offered in South African high schools. This gave the researchers a study sample 86, which reduced 81 following incomplete data submission by five students. The study’s findings suggest a sensitive understanding of students’ engagement with program materials and admission criteria. The admission criteria embraces prior high school knowledge reflected in the NSC-level admission points score (APS) and NSC-level mathematics. In addition, student dedication is a dimension of student engagement that we explored. The value of the study is a contribution to the reconsideration of student engagement alongside program admission requirements, leading to the integration of constructivist learning and student engagement theories. Prior studies have not integrated constructivist learning and student engagement theories into their research design. We recommend the continued use of the NSC-level APS score and mathematics in the admission requirements of the BAcc learning and development programs. We also recommend using teaching and learning activities that bring out student dedication. More research should be done on all aspects of student engagement premised on graduating students.

Extended Certainty Factors utilizing Neutrosophic Logic

Addressing uncertainty is a significant challenge in the field of artificial intelligence (AI). AI systems often encounter uncertainty due to incomplete data, ambiguous information, or inherent unpredictability in specific situations. To tackle this uncertainty, various solutions have been developed. These solutions range from probabilistic approaches like Bayesian networks and Monte Carlo methods to fuzzy logic and neural networks. These strategies enable AI systems to model and make judgments amid ambiguity by assigning probabilities, dealing with imprecise data, or utilizing learning processes that adapt to changing and uncertain contexts. However, probabilistic methods have limitations when it comes to handling uncertainty. These limitations include assumptions of independence, computational complexity, difficulty in capturing subjectivity, and interpretability. To address these limitations, other methods have been proposed, such as the model of certainty factors. This model offers a framework for reasoning under uncertainty by assigning a numerical value to statements or propositions. Additionally, neutrosophic logic extends classical logic by incorporating the concept of indeterminacy through truth/indeterminacy/falsity-membership functions. In this paper, we propose a method for investigating how certainty factors adapt to a neutrosophic environment. This method contributes to the development of more robust and adaptable decision support systems capable of dealing with diverse uncertainty. Our method, introduced for the first time in related literature, addresses issues such as the limited handling of indeterminacy, the inability to address contradictions, and the limited binary representation of uncertainty that characterize the model of certainty factors. We will provide a clear illustration of these implications through an example that demonstrates the superiority of our suggested method over the traditional technique of certainty factors.

Factors Associated with Health-Seeking Behavior in Indonesia: Evidence from the Indonesian Family Life Survey 2014

Background and Objectives: Health-seeking behavior is a critical determinant of health outcomes, particularly in countries like Indonesia. Given the increasing burden of noncommunicable diseases, understanding the factors that influence health-seeking behavior in this context is essential for developing more accessible and effective public health strategies. This study aimed to identify various factors associated with health-seeking behavior among patients with chronic diseases across Indonesia, especially in formal facilities. Materials and Methods: This study used a cross-sectional research design, utilizing Indonesian Family Life Survey (IFLS)-5 data. The inclusion criteria included respondents aged 20–74 years old with at least one chronic disease based on self-reported data. Respondents who did not receive treatment, practiced self-medication, or provided incomplete data were excluded. We used multivariate logistic regression to identify factors associated with health-seeking behavior in formal facilities. Results: The results revealed that 80.7% (n = 1993) of the 2471 respondents sought treatment in formal facilities, whereas 19.3% (n = 478) opted for informal facilities. Respondents who were Bugis (OR 9.187, 95% CI 2.182–38.683; p = 0.002), retired (OR 2.966, 95% CI 1.233–7.135; p = 0.015), did not smoke (OR 1.604, 95% CI 1.126–2.285; p = 0.009), made less than IDR 1,500,000 a month (OR 1.466, 95% CI 1.174–1.831; p = 0.000), had to travel more than 3 km to reach a treatment facility (OR 1.847, 95% CI 1.41–2.42; p = 0.000), or had more than one comorbidity (OR 1.396, 95% CI; p = 0.01) were more likely to seek treatment at formal facilities. Conclusions: These findings are expected to provide recommendations for policymakers, healthcare providers, and researchers to contribute to the development of targeted interventions that can improve healthcare access and utilization, ultimately enhancing health outcomes and equity in Indonesia.

Milost místo smrti (v procesech s cikány v Čechách první poloviny 18. století) aneb jak nepočítat historii

This article is concerned with the persecution of Gypsies in Bohemia in the first third of the 18th century. As in other European countries, in cases where Gypsies had been already formally expelled from a land, Gypsy vagabondage was defined and punished as a capital crime. The article does not forget this normative aspect of the theme, but it nonetheless concentrates on the actual practice of persecution and above all on cases in which condemned Gypsies begged for mercy and their death penalty was in fact reduced to a more moderate punishment. The author also looks at the extensive powers of reprieve that the Prague Apellate Court (in the case of Gypsy vagabondage the tribunal of first instance), was granted by the ruler in the 1720s. Condemned Gypsies were not explicitly mentioned in this context, but there is plenty of evidence that they were not excluded from this practice of reprieve. The article criticises the view of the persecution of Gypsies that is based solely on the quantification of incomplete data in the manuals of condemnation of the Apellate Court. These books not only fail to indicate when the condemned were later reprieved, but also do not allow us to reliably identify individuals condemned. There are examples of one person appearing several times in them, sometimes under different names. It is this misleading quantified evidence that has helped give rise to the idea that the persecution of Gypsies in the early modern period was the first stage of the Gypsy Holocaust ini the 20th century. The article argues, on the contrary, that in relation to persecution the pre-modern differed from the modern state not only in terms of capacity, but fundamentally.

Persian version of the Short Health Anxiety Inventory (SHAI) for patients with cancer: Evaluation of psychometric properties, factor structure, and association with related constructs

ObjectiveThis study aims to evaluate the psychometric properties of the Short Health Anxiety Inventory (SHAI=short health anxiety inventory) in a sample of patients diagnosed with cancer. Factor structure, internal consistency, convergent validity, and gender differences in SHAI scores were assessed. Methods202 patients diagnosed with cancer participated, with 18 questionnaires excluded for incomplete data. Confirmatory Factor Analysis was conducted to confirm the factor structure. Internal consistency was also assessed using Cronbach's alpha. ResultsA two-factor model (thought intrusion and fear of illness) provided the best fit for the data (SBχ2 (64) = 114.346, p < 0.001, CFI = 0.944, GFI = 0.908, RMSEA = 0.065). Internal consistency was high for the SHAI total score (α = 0.864) and its subscales (thought intrusion: α = 0.753; fear of illness: α = 0.825). Female patients also scored significantly higher than male patients on the SHAI total score (U = 5232.500, p = 0.012), thought intrusion subscale (U = 5189.00, p = 0.008) and fear of illness subscale (U = 5069.500, p = 0.018). ConclusionThe SHAI demonstrates adequate psychometric properties for assessing health anxiety in cancer patients.

Ship Trajectory Extraction Using Python Vessel Tracking Interpolation Method

A deep understanding of ship trajectory movements has essential implications for maritime applications, including navigation, monitoring, and marine traffic analysis. Efficient and accurate extraction techniques are necessary to extract valuable information from ship trajectory data. One commonly used approach is the interpolation method, which allows the reconstruction of smooth trajectories from recorded data points. This research is focused on analyzing the extraction of ship trajectories using the interpolation method provided by the Python Vessel Tracking (PyVT) library. This method allows interpolation of ship trajectory data based on various algorithms available in the library. This research aims to evaluate the effectiveness and accuracy of interpolation methods from PyVT in reconstructing ship trajectories from incomplete or disturbed data. Within this research's framework, several test scenarios were implemented to examine different types of ship trajectory data, including data with missing points and speed variations. Evaluation metrics RMSE (root mean Squared Error), which includes reconstruction accuracy, will be utilized from the interpolation algorithm in PyVT.

LSTM networks provide efficient cyanobacterial blooms forecasting even with incomplete spatio-temporal data

Cyanobacteria are the most frequent dominant species of algal blooms in inland waters, threatening ecosystem function and water quality, especially when toxin-producing strains predominate. Enhanced by anthropogenic activities and global warming, cyanobacterial blooms are expected to increase in frequency and global distribution. Early Warning Systems (EWS) for cyanobacterial blooms development allow timely implementation of management measures, reducing the risks associated to these blooms. In this paper, we propose an effective EWS for cyanobacterial bloom forecasting, which uses 6 years of incomplete high-frequency spatio-temporal data from multiparametric probes, including phycocyanin (PC) fluorescence as a proxy for cyanobacteria. A probe agnostic and replicable method is proposed to pre-process the data and to generate time series specific for cyanobacterial bloom forecasting. Using these pre-processed data, six different non-site/species-specific predictive models were compared including the autoregressive and multivariate versions of Linear Regression, Random Forest, and Long-Term Short-Term (LSTM) neural networks. Results were analyzed for seven forecasting time horizons ranging from 4 to 28 days evaluated with a hybrid system that combined regression metrics (MSE, R2, MAPE) for PC values, classification metrics (Accuracy, F1, Kappa) for a proposed alarm level of 10 µg PC/L, and a forecasting-specific metric to measure prediction improvement over the displaced signal (skill). The multivariate version of LSTM showed the best and most consistent results across all forecasting horizons and metrics, achieving accuracies of up to 90 % in predicting the proposed PC alarm level. Additionally, positive skill values indicated its outstanding effectiveness to forecast cyanobacterial blooms from 16 to 28 days in advance.

KAJIAN POTENSI INTERAKSI OBAT PADA PASIEN GERIATRI DENGAN GAGAL JANTUNG DI RSUD Dr. MOEWARDI SURAKARTA TAHUN 2023

Geriatric patients with heart failure generally receive more than 1 type of drug therapy, called polypharmacy, which typically involves at least four types of treatment: ACEI (Angiotensin Converting Enzyme Inhibitors), diuretics, beta blockers, and digoxin. This causes an increase in the occurrence of drug interactions. This study aims to determine the potential drug interactions that occur in geriatric patients with heart failure who undergo hospitalization at Dr. Moewardi Surakarta Hospital from January to December 2023. This study is a descriptive observational study with a cross-sectional design, and the data were collected retrospectively using a purposive sampling technique based on the medical records of geriatric patients at Dr. Moewardi Surakarta Hospital in 2023. The inclusion criteria were inpatients aged ≥55 years who were diagnosed with heart failure. The exclusion criteria were patients with incomplete or illegible medical record data or who died during treatment. Drug interaction research was conducted based on Drugs.com and Medscape. The results showed that out of the 100 patients, 98 experienced drug interactions. The incidence of drug interactions was 252 minor cases (21.65%), 818 moderate cases (69.50%), and 106 major cases (9.00%). The mechanism of drug interaction was pharmacodynamic in 919 cases (65.50%), pharmacokinetic in 213 cases (30.40%), and unknown in 44 cases (4.10%).

ТЕОРИЯ НЕЧЕТКИХ МНОЖЕСТВ В ИНТЕЛЛЕКТУАЛЬНОМ АНАЛИЗЕ ДАННЫХ

The article is devoted to the application of data mining (KDD) and fuzzy sets to work with large amounts of data. The digital revolution has led to a significant increase in the volume and speed of data, which makes traditional analysis methods ineffective. Data mining covers the stages of purification, integration, transformation and extraction of knowledge from data. Fuzzy logic, which extends Boolean logic, allows for uncertainty and approximate reasoning, which makes it especially useful for processing complex and noisy data. The article discusses in detail the applications of fuzzy sets in tasks such as clustering, associative rule detection, data generalization, time series analysis, and web applications. Fuzzy logic helps solve problems related to incomplete or inaccurate data by providing more flexible and understandable solutions. The author also notes that the use of fuzzy sets improves the adaptation of systems to specific data requirements and features, which contributes to more accurate analysis and decision-making in areas such as healthcare, finance and telecommunications. These advantages make fuzzy sets an important tool for improving the efficiency of data mining and improving user interaction.

Intelligent Design of Street Lamp in Rural Areas Based on an Improved Genetic Algorithm

This study addresses the demand for more efficient streetlight designs in rural areas by introducing an improved genetic algorithm (GA) to optimize the geometry and placement of streetlight poles. Conventional GAs frequently suffer from premature convergence and becoming trapped in local optima, reducing their effectiveness. To mitigate these issues, this research integrates the genetic algorithm with Sequential Quadratic Programming (SQP), using the quasi-optimal solution generated by the GA as the initial input for the SQP, enhancing both accuracy and stability. The methodology includes developing a geometric model of streetlight poles utilizing point cloud data and extracting the centerline via the optimized GA-SQP approach. Additionally, the study examines the effects of random errors, gross errors, incomplete point cloud data, and centerline deviations on the algorithm&amp;apos;s performance.

Computed tomography-based multi-organ radiomics nomogram model for predicting the risk of esophagogastric variceal bleeding in cirrhosis.

Radiomics has been used in the diagnosis of cirrhosis and prediction of its associated complications. However, most current studies predict the risk of esophageal variceal bleeding (EVB) based on image features at a single level, which results in incomplete data. Few studies have explored the use of global multi-organ radiomics for non-invasive prediction of EVB secondary to cirrhosis. To develop a model based on clinical and multi-organ radiomic features to predict the risk of first-instance secondary EVB in patients with cirrhosis. In this study, 208 patients with cirrhosis were retrospectively evaluated and randomly split into training (n = 145) and validation (n = 63) cohorts. Three areas were chosen as regions of interest for extraction of multi-organ radiomic features: The whole liver, whole spleen, and lower esophagus-gastric fundus region. In the training cohort, radiomic score (Rad-score) was created by screening radiomic features using the inter-observer and intra-observer correlation coefficients and the least absolute shrinkage and selection operator method. Independent clinical risk factors were selected using multivariate logistic regression analyses. The radiomic features and clinical risk variables were combined to create a new radiomics-clinical model (RC model). The established models were validated using the validation cohort. The RC model yielded the best predictive performance and accurately predicted the EVB risk of patients with cirrhosis. Ascites, portal vein thrombosis, and plasma prothrombin time were identified as independent clinical risk factors. The area under the receiver operating characteristic curve (AUC) values for the RC model, Rad-score (liver + spleen + esophagus), Rad-score (liver), Rad-score (spleen), Rad-score (esophagus), and clinical model in the training cohort were 0.951, 0.930, 0.801, 0.831, 0.864, and 0.727, respectively. The corresponding AUC values in the validation cohort were 0.930, 0.886, 0.763, 0.792, 0.857, and 0.692. In patients with cirrhosis, combined multi-organ radiomics and clinical model can be used to non-invasively predict the probability of the first secondary EVB.

Characterization of myeloproliferative neoplasms based on genetics only and prognostication of transformation to blast phase.

Myeloproliferative neoplasms (MPN) are a heterogeneous group of clonal disorders characterized by aberrant hematopoietic proliferation and an intrinsic risk of progression to blast phase. The WHO classification 2022 identifies chronic myeloid leukemia and the BCR::ABL1 negative MPNs polycythemia vera, primary myelofibrosis and essential thrombocythemia as individual entities. However, overlaps, borderline findings or transitions between MPN subtypes occur and incomplete clinical data often complicates diagnosis. By conducting a thorough genetic analysis, we've developed a model that relies on 12 genetic markers to accurately stratify MPN patients. The model can be simplified into a decision tree for routine use. Comparing samples at chronic and blast phase revealed, that one third of patients lost their MPN driver-gene mutation, while mutations in splicing and chromatin modifying genes were stable, indicating a shared founder clone of chronic and blast phase with different driver mutations and therefore different progressing capacities. This was further supported by gain of typical de novo AML gene mutations, accompanied by gain of complex karyotypes and RAS pathway gene mutations. Our data suggest to perform a broader genetic screening at diagnosis and also at clinical progression, as driver mutations may change and the MPN-driver mutations present at diagnosis may disappear.

Development and optimization of an Artificial Neural Network (ANN) model for predicting the cadmium fixation efficiency of biochar in soil

This study addresses the issue of soil cadmium pollution and the common problem of data missing during the training of artificial neural network models. A substantial amount of characteristic data regarding the immobilization of cadmium in soil through biochar was collected from the literature. Subsequently, data was organized and analyzed using correlation matrix analysis and feature importance analysis. The results revealed that the feature importance of biochar (52%) was relatively significant. A novel approach is introduced, involving a non-deterministic optimization technique based on linear interpolation to supplement the missing data within the dataset. The completed dataset effectively preserves the distributional characteristics of the original data, thereby enhancing the training performance of the neural network model. Through three phases, namely data incompleteness, data supplementation, and genetic algorithm optimization, the neural network model was progressively refined. The outcome of this iterative process is a high-performance artificial neural network model capable of predicting the efficiency of biochar in immobilizing cadmium in soil. The significance of these findings lies in their potential to advance the application of machine learning in environmental science, particularly in addressing complex pollution scenarios.

Multiobjective optimization guided by image quality index for limited-angle CT image reconstruction.

Due to the incomplete projection data collected by limited-angle computed tomography (CT), severe artifacts are present in the reconstructed image. Classical regularization methods such as total variation (TV) minimization, ℓ0 minimization, are unable to suppress artifacts at the edges perfectly. Most existing regularization methods are single-objective optimization approaches, stemming from scalarization methods for multiobjective optimization problems (MOP). To further suppress the artifacts and effectively preserve the edge structures of the reconstructed image. This study presents a multiobjective optimization model incorporates both data fidelity term and ℓ0-norm of the image gradient as objective functions. It employs an iterative approach different from traditional scalarization methods, using the maximization of structural similarity (SSIM) values to guide optimization rather than minimizing the objective function.The iterative method involves two steps, firstly, simultaneous algebraic reconstruction technique (SART) optimizes the data fidelity term using SSIM and the Simulated Annealing (SA) algorithm for guidance. The degradation solution is accepted in the form of probability, and guided image filtering (GIF) is introduced to further preserve the image edge when the degradation solution is rejected. Secondly, the result from the first step is integrated into the second objective function as a constraint, we use ℓ0 minimization to optimize ℓ0-norm of the image gradient, and the SSIM, SA algorithm and GIF are introduced to guide optimization process by improving SSIM value like the first step. With visual inspection, the peak signal-to-noise ratio (PSNR), root mean square error (RMSE), and SSIM values indicate that our approach outperforms other traditional methods. The experiments demonstrate the effectiveness of our method and its superiority over other classical methods in artifact suppression and edge detail restoration.

Multi-View Integrative Approach For Imputing Short-Chain Fatty Acids and Identifying Key factors predicting Blood SCFA.

Short-chain fatty acids (SCFAs) are the main metabolites produced by bacterial fermentation of dietary fiber within gastrointestinal tract. SCFAs produced by gut microbiotas (GMs) are absorbed by host, reach bloodstream, and are distributed to different organs, thus influencing host physiology. However, due to the limited budget or the poor sensitivity of instruments, most studies on GMs have incomplete blood SCFA data, limiting our understanding of the metabolic processes within the host. To address this gap, we developed an innovative multi-task multi-view integrative approach (M2AE, Multi-task Multi-View Attentive Encoders), to impute blood SCFA levels using gut metagenomic sequencing (MGS) data, while taking into account the intricate interplay among the gut microbiome, dietary features, and host characteristics, as well as the nuanced nature of SCFA dynamics within the body. Here, each view represents a distinct type of data input (i.e., gut microbiome compositions, dietary features, or host characteristics). Our method jointly explores both view-specific representations and cross-view correlations for effective predictions of SCFAs. We applied M2AE to two in-house datasets, which both include MGS and blood SCFAs profiles, host characteristics, and dietary features from 964 subjects and 171 subjects, respectively. Results from both of two datasets demonstrated that M2AE outperforms traditional regression-based and neural-network based approaches in imputing blood SCFAs. Furthermore, a series of gut bacterial species (e.g., Bacteroides thetaiotaomicron and Clostridium asparagiforme), host characteristics (e.g., race, gender), as well as dietary features (e.g., intake of fruits, pickles) were shown to contribute greatly to imputation of blood SCFAs. These findings demonstrated that GMs, dietary features and host characteristics might contribute to the complex biological processes involved in blood SCFA productions. These might pave the way for a deeper and more nuanced comprehension of how these factors impact human health.

Empowering Graph Neural Network-Based Computational Drug Repositioning with Large Language Model-Inferred Knowledge Representation.

Computational drug repositioning, through predicting drug-disease associations (DDA), offers significant potential for discovering new drug indications. Current methods incorporate graph neural networks (GNN) on drug-disease heterogeneous networks to predict DDAs, achieving notable performances compared to traditional machine learning and matrix factorization approaches. However, these methods depend heavily on network topology, hampered by incomplete and noisy network data, and overlook the wealth of biomedical knowledge available. Correspondingly, large language models (LLMs) excel in graph search and relational reasoning, which can possibly enhance the integration of comprehensive biomedical knowledge into drug and disease profiles. In this study, we first investigate the contribution of LLM-inferred knowledge representation in drug repositioning and DDA prediction. A zero-shot prompting template was designed for LLM to extract high-quality knowledge descriptions for drug and disease entities, followed by embedding generation from language models to transform the discrete text to continual numerical representation. Then, we proposed LLM-DDA with three different model architectures (LLM-DDANode Feat, LLM-DDADual GNN, LLM-DDAGNN-AE) to investigate the best fusion mode for LLM-based embeddings. Extensive experiments on four DDA benchmarks show that, LLM-DDAGNN-AE achieved the optimal performance compared to 11 baselines with the overall relative improvement in AUPR of 23.22%, F1-Score of 17.20%, and precision of 25.35%. Meanwhile, selected case studies of involving Prednisone and Allergic Rhinitis highlighted the model's capability to identify reliable DDAs and knowledge descriptions, supported by existing literature. This study showcases the utility of LLMs in drug repositioning with its generality and applicability in other biomedical relation prediction tasks.