Enhanced knowledge graph recommendation algorithm based on multi-level contrastive learning.

Integrating the Knowledge Graphs (KGs) into recommendation systems enhances personalization and accuracy. However, the long-tail distribution of knowledge graphs often leads to data sparsity, which limits the effectiveness in practical applications. To address this challenge, this study proposes a knowledge-aware recommendation algorithm framework that incorporates multi-level contrastive learning. This framework enhances the Collaborative Knowledge Graph (CKG) through a random edge dropout method, which constructs feature representations at three levels: user-user interactions, item-item interactions and user-item interactions. A dynamic attention mechanism is employed in the Graph Attention Networks (GAT) for modeling the KG. Combined with the nonlinear transformation and Momentum Contrast (Moco) strategy for contrastive learning, it can effectively extract high-quality feature information. Additionally, multi-level contrastive learning, as an auxiliary self-supervised task, is jointly trained with the primary supervised task, which further enhances recommendation performance. Experimental results on the MovieLens and Amazon-books datasets demonstrate that this framework effectively improves the performance of knowledge graph-based recommendations, addresses the issue of data sparsity, and outperforms other baseline models across multiple evaluation metrics.

High-resolution lensless holographic microscopy using a physics-aware deep network.

Significance: Lensless digital inline holographic microscopy (LDIHM) is an emerging quantitative phase imaging modality that uses advanced computational methods for phase retrieval from the interference pattern. The existing end-to-end deep networks require a large training dataset with sufficient diversity to achieve high-fidelity hologram reconstruction. To mitigate this data requirement problem, physics-aware deep networks integrate the physics of holography in the loss function to reconstruct complex objects without needing prior training. However, the data fidelity term measures the data consistency with a single low-resolution hologram without any external regularization, which results in a low performance on complex biological data. Aim: We aim to mitigate the challenges with trained and physics-aware untrained deep networks separately and combine the benefits of both methods for high-resolution phase recovery from a single low-resolution hologram in LDIHM. Approach: We propose a hybrid deep framework (HDPhysNet) using a plug-and-play method that blends the benefits of trained and untrained deep models for phase recovery in LDIHM. The high-resolution phase is generated by a pre-trained high-definition generative adversarial network (HDGAN) from a single low-resolution hologram. The generated phase is then plugged into the loss function of a physics-aware untrained deep network to regulate the complex object reconstruction process. Results: Simulation results show that the SSIM of the proposed method is increased by 0.07 over the trained and 0.04 over the untrained deep networks. The average phase-SNR is elevated by 8.2 dB over trained deep models and 9.8 dB over untrained deep networks on the experimental biological cells (cervical cells and red blood cells). Conclusions: We showed improved performance of the HDPhysNet against the unknown perturbation in the imaging parameters such as the propagation distance, the wavelength of the illuminating source, and the imaging sample compared with the trained network (HDGAN). LDIHM, combined with HDPhysNet, is a portable and technology-driven microscopy best suited for point-of-care cytology applications.

Antimicrobial Use Among Outpatients at Benjamin Mkapa Hospital in Dodoma Central Zone of Tanzania: A Prospective Descriptive Study.

Objective: Irrational antimicrobial use (AMU) has led to an exponential increase in antimicrobial resistance (AMR) in hospitals and communities, which creates challenges in treating infectious diseases caused by bacteria. This study aimed to evaluate antimicrobial prescriptions and usage patterns for treating bacterial infections among outpatients at Benjamin Mkapa Hospital (BMH). Materials and Methods: A prospective descriptive study design was used to evaluate the AMU trend. The data were collected from August 2022 to October 2022 from outpatient pharmacies at BMH using the World Health Organization/International Network of Rational Use of Drugs (WHO/INRUD) indicators. The simple random sampling method was employed to select the prescriptions. The WHO AWaRe (Access, Watch, and Reserve) classification was used to classify common antimicrobials. We analyzed the prevalence of outpatient AMU, including the types of antimicrobials, indications, and compliance with treatment guidelines. We also examined the number of antimicrobials per prescription and the adherence to drug use. Results: We examined 1557 prescriptions, 406 (26.1%) (WHO recommendation 20.0-26.8%) of which included antimicrobials. All prescriptions with antimicrobials were written in generic names, drug utilization-90% (DU90%) was 100% (WHO recommendation 100%). The number of parenteral antimicrobials prescribed was 79 (19.5%) (WHO recommendation 13.4-24.1%). Furthermore, prescriptions with antimicrobials that complied with the current Standard Treatment Guidelines and National Essential Medicine List in Tanzania (STG/NEMLIT) were 369 (90.9%) (WHO recommendation 100%). Most antimicrobials were prescribed as monotherapy, accounting for 265 (65.3%). There were 1.4 (WHO recommendation 1.6-1.8) antimicrobials per prescription. Our study identified 21 commonly prescribed antimicrobials, whereby 9 (42.9%) (WHO recommendation >60%) antimicrobials were Access, 10 (47.6%) (WHO recommendation <20%) Watch, and 2 (9.5%) (WHO recommendation <1%) Reserved classes. Conclusion: Our study showed that BMH has optimal practices for prescribing and using antimicrobials for outpatients. It further underlined the need to expand and strengthen antimicrobial stewardship efforts to reinforce prescribing antimicrobials.

Abnormal recognition-assisted and onset-offset aware network for pathological wearable ECG delineation.

Electrocardiogram (ECG) delineation is essential to the identification of abnormal cardiac status, especially when ECG signals are remotely monitored with wearable devices. The complexity and diversity of cardiac conditions generate numerous pathological ECG patterns, not only requiring the recognition of normal ECG but also addressing an extensive range of abnormal ECG patterns, posing a challenging task. Therefore, we propose an abnormal recognition-assisted network to integrate supplementary information on diverse ECG patterns. Simultaneously, we design an onset-offset aware loss to enhance precise waveform localization. Specifically, we establish a two-branch framework where ECG delineation serves as the target task, producing the final segmentation results. Additionally, the abnormal recognition-assisted network serves as an auxiliary task, extracting multi-label pathological information from ECGs. This joint learning approach establishes crucial correlations between ECG delineation and associated ECG abnormalities. The correlations enable the model to demonstrate sufficient generalization in the presence of diverse abnormal ECG patterns. Besides, onset-offset aware loss focuses intensively on wave onsets and offsets by applying biased weights to various waveform positions. This approach ensures a focus on precise localization, facilitating seamless integration into cross-entropy loss function. A large-scale wearable 12-lead dataset containing 4,913 signals is collected, offering an extensive range of ECG data for model training. Results demonstrate that our method achieves outstanding performance on two test datasets, attaining sensitivity of 94.97% and 94.27% and an error tolerance lower than 20 ms. Furthermore, our method is effective for various aberrant ECG signals, including ST-segment changes, atrial premature beats, and right and left bundle branch blocks.

Antibiotic prescribing patterns in pediatric patients using the WHO access, watch, reserve (AWaRe) classification at a quaternary hospital in Nampula, Mozambique.

Antibiotics are often prescribed inappropriately, either when they are not needed or with an unnecessarily broad spectrum of activity. This is a serious problem that can lead to the development of antimicrobial resistance (AMR). This study was conducted to assess the antibiotic prescribing pattern in pediatric patients hospitalized at a quaternary hospital in Nampula, Mozambique, using the WHO indicators and Framework as a reference. A retrospective study was conducted using secondary data obtained from medical records. The study population consisted of children aged 0-10 years who were hospitalized in a quaternary-level hospital ward in Nampula, Mozambique. The pattern of antibiotic prescriptions was assessed using indicators and the WHO classification of antibiotics into AWaRe categories. Descriptive statistics were applied. A total of 464 antibiotics were prescribed during the study. The age groups of 1-3 years and 28 days-12 months were prescribed more antibiotics. The most common antibiotics were ceftriaxone and crystallized penicillin, which were frequently prescribed for patients suffering from bronchopneumonia, gastroenteritis, and malaria. 74.8% of the antibiotics prescribed belonged to the Access group, while 23.7% belonged to the Watch group. There were no prescriptions of antibiotics from the Reserve group. The average number of antibiotics per prescription was 1.51 (SD ± 0.725). The percentage of antibiotic prescribing was 97.5%, with 96.20% by injection. All antibiotics prescribed were on the essential medicines list and prescribed by generic name. These results are concerning and highlight the urgency of strengthening antimicrobial optimization measures, as well as implementing the AWaRe framework in antibiotic prescribing as an essential strategy to combat AMR.

Returned Rate and Changed Patterns of Systemic Antibiotic Use in Ambulatory Care in Hungary after the Pandemic-A Longitudinal Ecological Study.

The COVID-19 pandemic affected the epidemiology of infectious diseases and changed the operation of health care systems and health care seeking behavior. Our study aimed to analyze the utilization of systemic antibiotics in ambulatory care in Hungary after the COVID-19 pandemic and compare it to the period before COVID. We defined three periods (24 months each): Before COVID, COVID, and After COVID. Monthly trends in systemic antibiotic (J01) use were calculated using the WHO ATC-DDD index and expressed as DDD/1000 inhabitants/day (DID) and number of exposed patients per active agent. The data were further categorized by the WHO AWaRe classification. In the After COVID period, we detected almost the same (11.61 vs. 11.11 DID) mean monthly use of systemic antibiotics in ambulatory care compared to the Before COVID period. We observed a decrease in the seasonality index in the After COVID period (46.86% vs. 39.86%). In the After COVID period, the use of cephalosporins and quinolones decreased significantly, while in the case of macrolides, a significant increase was observed compared to the Before COVID period, with excessive azithromycin use (66,869 vs. 97,367 exposed patients). This study demonstrated significant changes in the pattern of ambulatory care antibiotic use in Hungary.

A Lesion-aware Edge-based Graph Neural Network for Predicting Language Ability in Patients with Post-stroke Aphasia.

We propose a lesion-aware graph neural network (LEGNet) to predict language ability from resting-state fMRI (rs-fMRI) connectivity in patients with post-stroke aphasia. Our model integrates three components: an edge-based learning module that encodes functional connectivity between brain regions, a lesion encoding module, and a subgraph learning module that leverages functional similarities for prediction. We use synthetic data derived from the Human Connectome Project (HCP) for hyperparameter tuning and model pretraining. We then evaluate the performance using repeated 10-fold cross-validation on an in-house neuroimaging dataset of post-stroke aphasia. Our results demonstrate that LEGNet outperforms baseline deep learning methods in predicting language ability. LEGNet also exhibits superior generalization ability when tested on a second in-house dataset that was acquired under a slightly different neuroimaging protocol. Taken together, the results of this study highlight the potential of LEGNet in effectively learning the relationships between rs-fMRI connectivity and language ability in a patient cohort with brain lesions for improved post-stroke aphasia evaluation.

DualAttlog: Context aware dual attention networks for log-based anomaly detection.

Most existing log-driven anomaly detection methods assume that logs are static and unchanged, which is often impractical. To address this, we propose a log anomaly detection model called DualAttlog. This model includes word-level and sequence-level semantic encoding modules, as well as a context-aware dual attention module. Specifically, The word-level semantic encoding module utilizes a self-matching attention mechanism to explore the interactive properties between words in log sequences. By performing word embedding and semantic encoding, it captures the associations and evolution processes between words, extracting local-level semantic information. while The sequence-level semantic encoding module encoding the entire log sequence using a pre-trained model. This extracts global semantic information, capturing overall patterns and trends in the logs. The context-aware dual attention module integrates these two levels of encoding, utilizing contextual information to reduce redundancy and enhance detection accuracy. Experimental results show that the DualAttlog model achieves an F1-Score of over 95% on 7 public datasets. Impressively, it achieves an F1-Score of 82.35% on the Real-Industrial W dataset and 83.54% on the Real-Industrial Q dataset. It outperforms existing baseline techniques on 9 datasets, demonstrating its significant advantages.

The skinny on skin: The role of skin-aware professionals in skin cancer surveillance.

BACKGROUND: Licensed nonmedical, skin-aware professionals (e.g., hairdressers, massage therapists, etc.) have the potential to identify skin cancer, but baseline knowledge may not be sufficient to accomplish this goal. Following educational intervention, self-efficacy is one of the best surrogate metrics for behavior change. Curricula that increase knowledge and confidence levels can improve screening behaviors, but few have been tested for efficacy in this population AIMS: We assessed whether an online curriculum could reliably improve skin screening knowledge, attitudes, and behaviors of nonmedical professionals PATIENTS/METHODS: Skin-aware professionals were recruited through the Oregon Health Authority and IMPACT Melanoma TM. Participants completed a pre-survey, online training module, post-survey, and one-year follow-up survey. We evaluated participants' indicated levels of concern for suspicious and nonsuspicious lesions relative to "gold standard" physician ratings. We also assessed confidence and self-reported behavior change regarding talking to clients about skin cancer and recommending they see a provider to evaluate suspicious lesions RESULTS: The pre-survey was completed by 9872 skin-aware professionals; 5434 completed the post-survey, and 162 completed the one-year follow-up survey. Participants showed a significant improvement in ability to indicate the correct level of concern for all lesion types in concordance with "gold standard" physician ratings (p < 0.001). Participants reported increased comfort levels in discussing health-related topics with their clients posttraining CONCLUSIONS: Our training module effectively increased skin-aware professionals' knowledge, confidence, and concern for malignant lesions. Skin-aware professionals may serve as a valuable extension of the skin self-exam, but additional studies are needed to evaluate the impact of these curricula long-term, including potential downstream consequences.

Data privacy-aware machine learning approach in pancreatic cancer diagnosis.

PROBLEM: Pancreatic ductal adenocarcinoma (PDAC) is considered a highly lethal cancer due to its advanced stage diagnosis. The five-year survival rate after diagnosis is less than 10%. However, if diagnosed early, the five-year survival rate can reach up to 70%. Early diagnosis of PDAC can aid treatment and improve survival rates by taking necessary precautions. The challenge is to develop a reliable, data privacy-aware machine learning approach that can accurately diagnose pancreatic cancer with biomarkers. AIM: The study aims to diagnose a patient's pancreatic cancer while ensuring the confidentiality of patient records. In addition, the study aims to guide researchers and clinicians in developing innovative methods for diagnosing pancreatic cancer. METHODS: Machine learning, a branch of artificial intelligence, can identify patterns by analyzing large datasets. The study pre-processed a dataset containing urine biomarkers with operations such as filling in missing values, cleaning outliers, and feature selection. The data was encrypted using the Fernet encryption algorithm to ensure confidentiality. Ten separate machine learning models were applied to predict individuals with PDAC. Performance metrics such as F1 score, recall, precision, and accuracy were used in the modeling process. RESULTS: Among the 590 clinical records analyzed, 199 (33.7%) belonged to patients with pancreatic cancer, 208 (35.3%) to patients with non-cancerous pancreatic disorders (such as benign hepatobiliary disease), and 183 (31%) to healthy individuals. The LGBM algorithm showed the highest efficiency by achieving an accuracy of 98.8%. The accuracy of the other algorithms ranged from 98 to 86%. In order to understand which features are more critical and which data the model is based on, the analysis found that the features "plasma_CA19_9", REG1A, TFF1, and LYVE1 have high importance levels. The LIME analysis also analyzed which features of the model are important in the decision-making process. CONCLUSIONS: This research outlines a data privacy-aware machine learning tool for predicting PDAC. The results show that a promising approach can be presented for clinical application. Future research should expand the dataset and focus on validation by applying it to various populations.

Psychosis Prognosis Predictor: A continuous and uncertainty-aware prediction of treatment outcome in first-episode psychosis.

INTRODUCTION: Machine learning models have shown promising potential in individual-level outcome prediction for patients with psychosis, but also have several limitations. To address some of these limitations, we present a model that predicts multiple outcomes, based on longitudinal patient data, while integrating prediction uncertainty to facilitate more reliable clinical decision-making. MATERIAL AND METHODS: We devised a recurrent neural network architecture incorporating long short-term memory (LSTM) units to facilitate outcome prediction by leveraging multimodal baseline variables and clinical data collected at multiple time points. To account for model uncertainty, we employed a novel fuzzy logic approach to integrate the level of uncertainty into individual predictions. We predicted antipsychotic treatment outcomes in 446 first-episode psychosis patients in the OPTiMiSE study, for six different clinical scenarios. The treatment outcome measures assessed at both week 4 and week 10 encompassed symptomatic remission, clinical global remission, and functional remission. RESULTS: Using only baseline predictors to predict different outcomes at week 4, leave-one-site-out validation AUC ranged from 0.62 to 0.66; performance improved when clinical data from week 1 was added (AUC = 0.66-0.71). For outcome at week 10, using only baseline variables, the models achieved AUC = 0.56-0.64; using data from more time points (weeks 1, 4, and 6) improved the performance to AUC = 0.72-0.74. After incorporating prediction uncertainties and stratifying the model decisions based on model confidence, we could achieve accuracies above 0.8 for ~50% of patients in five out of the six clinical scenarios. CONCLUSION: We constructed prediction models utilizing a recurrent neural network architecture tailored to clinical scenarios derived from a time series dataset. One crucial aspect we incorporated was the consideration of uncertainty in individual predictions, which enhances the reliability of decision-making based on the model's output. We provided evidence showcasing the significance of leveraging time series data for achieving more accurate treatment outcome prediction in the field of psychiatry.

RDSC: Range-Based Device Spatial Clustering for IoT Networks.

The growth of the Internet of Things (IoT) has become a crucial area of modern research. While the increasing number of IoT devices has driven significant advancements, it has also introduced several challenges, such as data storage, data privacy, communication protocols, complex network topologies, and IoT device management. In essence, the management of IoT devices is becoming more and more challenging, especially with the limited capacity and power of the IoT devices. The devices, having limited capacities, cannot store the information of the entire environment at once. In addition, device power consumption can affect network performance and stability. The devices' sensing areas with device grouping and management can simplify further networking tasks and improve response quality with data aggregation and correction techniques. In fact, most research papers are looking forward to expanding network lifetimes by relying on devices with high power capabilities. This paper proposes a device spatial clustering technique that covers crucial challenges in IoT. Our approach groups the dispersed devices to create clusters of connected devices while considering their coverage, their storage capacities, and their power. A new clustering protocol alongside a new clustering algorithm is introduced, resolving the aforementioned challenges. Moreover, a technique for non-sensed area extraction is presented. The efficiency of the proposed approach has been evaluated with extensive experiments that gave notable results. Our technique was also compared with other clustering algorithms, showing the different results of these algorithms.

Improving effort-aware just-in-time defect prediction with weighted code churn and multi-objective slime mold algorithm.

Effort-aware just-in-time software defect prediction (JIT-SDP) aims to effectively utilize the limited resources of software quality assurance (SQA) to detect more software defects. This improves the efficiency of SQA work and the quality of the software. However, there is disagreement regarding the representation of the key feature variable, SQA effort, in the field of effort-aware JIT-SDP. Additionally, the most recent metaheuristic optimization algorithms (MOAs) have not yet been effectively integrated with multi-objective effort-aware JIT-SDP tasks. These deficiencies, in both feature representation and model optimization (MO), result in a significant disparity between the performance of effort-aware JIT-SDP techniques and the expectations of the industry. In this study, we present a novel method called weighted code churn (CC) and improved multi-objective slime mold algorithm (SMA) (WCMS) for effort-aware JIT-SDP. It comprises two stages: feature improvement (FI) and MO. In the FI phase, we normalize the two feature variables: number of modified files (NF) and distribution of modified code across each file (Entropy). We then use an exponential function to quantify the level of difficulty of the change. The equation is as follows: DD = NFEntropy, where DD is an acronym for the degree of difficulty, NF denotes the base number, and Entropy denotes the index. We define change effort as the product of the difficulty degree in implementing the change and CC, with weighted CC representing the change effort. During the MO stage, we improve the SMA by incorporating multi-objective handling capabilities and devising mechanisms for multi-objective synchronization and conflict resolution. We develop a multi-objective optimization algorithm for hyperparameter optimization (HPO) of the JIT-SDP model in WCMS. To evaluate the performance of our method, we conducted experiments using six well-known open-source projects and employed two effort-aware performance evaluation metrics. We evaluated our method based on three scenarios: cross-validation, time-wise cross-validation, and across-project prediction. The experimental results indicate that the proposed method outperforms the benchmark method. Furthermore, the proposed method demonstrates greater scalability and generalization capabilities.

Understanding antibiotic purchasing practices in community pharmacies: A potential driver of emerging antimicrobial resistance.

Introduction: Antimicrobial resistance (AMR), a transboundary health issue, critically impacting low- and middle-income countries (LMICs) where 80% of antibiotics are used in the community, with 20-50% being inappropriate. Southeast-Asia, including Bangladesh, faces heightened AMR risk due to suboptimal healthcare standard and unregulated antibiotic sales. This study aimed to audit antibiotic dispensing patterns from community pharmacies, identifying factors influencing purchasing behaviors. Methods: A cross-sectional survey of 385 antibiotic customers and structured observations of 1000 pharmacy dispensing events were conducted in four urban and rural areas in Bangladesh. Descriptive analysis defined antibiotic use, while Poisson regression examined how patients' demographics and health symptoms influenced prescription behaviors. Results: Among 1000 observed medicine dispensing events, 25.9% were antibiotics. Commonly purchased antibiotics included macrolides (22.8%), third-generation-cephalosporins (20.8%), and second-generation-cephalosporins (16.9%). Following WHO-AWaRe classifications, 73.5% of antibiotics were categorized as Watch, and 23.1% as Access. From the survey, 56.6% antibiotics were purchased without a prescription from drug-sellers and informal healthcare providers, primarily for "non-severe" health-symptoms such as upper-respiratory-tract infections (37.4%), fever (31.7%), uncomplicated skin infections (20%), gastrointestinal-infections (11.2%), and urinary-tract infections (7.9%). The likelihood of presenting a prescription while purchasing antibiotics was 27% lower for individuals aged 6-59 compared to those ≤5 or ≥ 60. Lower-respiratory-tract infections and enteric-fever had higher prescription rates, with adjusted prevalence ratios of 1.78 (95% CI: 1.04, 3.03) and 1.87 (95% CI: 1.07, 3.29), respectively. After adjusting for confounders, sex, urban-rural locations, income, education, and number of health-symptoms exhibited no significant influence on prescription likelihood. Conclusion: This study underscores unregulated antibiotic sales without prescriptions, urging tailored interventions considering prevailing health-seeking practices in diverse healthcare settings in LMICs. Enforcing prescription-only regulations is hindered by easy access through community pharmacies and conflicts of interest. Future strategies should consider how stewardship impacts the financial interests of pharmacy personnel in settings lacking clear authority to ensure optimal compliance.

Knowledge graph confidence-aware embedding for recommendation.

Knowledge graphs (KG) are vital for extracting and storing knowledge from large datasets. Current research favors knowledge graph-based recommendation methods, but they often overlook the features learning of relations between entities and focus excessively on entity-level details. Moreover, they ignore a crucial fact: the aggregation process of entity and relation features in KG is complex, diverse, and imbalanced. To address this, we propose a recommendation-oriented KG confidence-aware embedding technique. It introduces an information aggregation graph and a confidence feature aggregation mechanism to overcome these challenges. Additionally, we quantify entity confidence at the feature and category levels, improving the precision of embeddings during information propagation and aggregation. Our approach achieves significant improvements over state-of-the-art KG embedding-based recommendation methods, with up to 6.20% increase in AUC and 8.46% increase in GAUC, as demonstrated on four public KG datasets2.

DART: DEFORMABLE ANATOMY-AWARE REGISTRATION TOOLKIT FOR LUNG CT REGISTRATION WITH KEYPOINTS SUPERVISION.

Spatially aligning two computed tomography (CT) scans of the lung using automated image registration techniques is a challenging task due to the deformable nature of the lung. However, existing deep-learning-based lung CT registration models are not trained with explicit anatomical knowledge. We propose the deformable anatomy-aware registration toolkit (DART), a masked autoencoder (MAE)-based approach, to improve the keypoint-supervised registration of lung CTs. Our method incorporates features from multiple decoders of networks trained to segment anatomical structures, including the lung, ribs, vertebrae, lobes, vessels, and airways, to ensure that the MAE learns relevant features corresponding to the anatomy of the lung. The pretrained weights of the transformer encoder and patch embeddings are then used as the initialization for the training of downstream registration. We compare DART to existing state-of-the-art registration models. Our experiments show that DART outperforms the baseline models (Voxelmorph, ViT-V-Net, and MAE-TransRNet) in terms of target registration error of both corrField-generated keypoints with 17%, 13%, and 9% relative improvement, respectively, and bounding box centers of nodules with 27%, 10%, and 4% relative improvement, respectively. Our implementation is available at https://github.com/yunzhengzhu/DART.

Optimizing agricultural data security: harnessing IoT and AI with Latency Aware Accuracy Index (LAAI).

The integration of Internet of Things (IoT) and artificial intelligence (AI) technologies into modern agriculture has profound implications on data collection, management, and decision-making processes. However, ensuring the security of agricultural data has consistently posed a significant challenge. This study presents a novel evaluation metric titled Latency Aware Accuracy Index (LAAI) for the purpose of optimizing data security in the agricultural sector. The LAAI uses the combined capacities of the IoT and AI in addition to the latency aspect. The use of IoT tools for data collection and AI algorithms for analysis makes farming operation more productive. The LAAI metric is a more holistic way to determine data accuracy while considering latency limitations. This ensures that farmers and other end-users are fed trustworthy information in a timely manner. This unified measure not only makes the data more secure but gives farmers the information that helps them to make smart decisions and, thus, drives healthier farming and food security.

Empathic Conversational Agent Platform Designs and Their Evaluation in the Context of Mental Health: Systematic Review.

BACKGROUND: The demand for mental health (MH) services in the community continues to exceed supply. At the same time, technological developments make the use of artificial intelligence-empowered conversational agents (CAs) a real possibility to help fill this gap. OBJECTIVE: The objective of this review was to identify existing empathic CA design architectures within the MH care sector and to assess their technical performance in detecting and responding to user emotions in terms of classification accuracy. In addition, the approaches used to evaluate empathic CAs within the MH care sector in terms of their acceptability to users were considered. Finally, this review aimed to identify limitations and future directions for empathic CAs in MH care. METHODS: A systematic literature search was conducted across 6 academic databases to identify journal articles and conference proceedings using search terms covering 3 topics: "conversational agents," "mental health," and "empathy." Only studies discussing CA interventions for the MH care domain were eligible for this review, with both textual and vocal characteristics considered as possible data inputs. Quality was assessed using appropriate risk of bias and quality tools. RESULTS: A total of 19 articles met all inclusion criteria. Most (12/19, 63%) of these empathic CA designs in MH care were machine learning (ML) based, with 26% (5/19) hybrid engines and 11% (2/19) rule-based systems. Among the ML-based CAs, 47% (9/19) used neural networks, with transformer-based architectures being well represented (7/19, 37%). The remaining 16% (3/19) of the ML models were unspecified. Technical assessments of these CAs focused on response accuracies and their ability to recognize, predict, and classify user emotions. While single-engine CAs demonstrated good accuracy, the hybrid engines achieved higher accuracy and provided more nuanced responses. Of the 19 studies, human evaluations were conducted in 16 (84%), with only 5 (26%) focusing directly on the CA's empathic features. All these papers used self-reports for measuring empathy, including single or multiple (scale) ratings or qualitative feedback from in-depth interviews. Only 1 (5%) paper included evaluations by both CA users and experts, adding more value to the process. CONCLUSIONS: The integration of CA design and its evaluation is crucial to produce empathic CAs. Future studies should focus on using a clear definition of empathy and standardized scales for empathy measurement, ideally including expert assessment. In addition, the diversity in measures used for technical assessment and evaluation poses a challenge for comparing CA performances, which future research should also address. However, CAs with good technical and empathic performance are already available to users of MH care services, showing promise for new applications, such as helpline services.

Investigating the WHO's AWaRe classification for assessing antimicrobial stewardship programs: A single-center study.

BACKGROUND: AWaRe (Access, Watch, Reserve) classification proposed by the World Health Organization (WHO) holds potential for assessing antimicrobial stewardship programs (ASPs). However, increase in antibiotics for non-infectious treatment might undermine the effectiveness of using the AWaRe classification for assessing ASPs. The study aimed to evaluate the antimicrobial usage by AWaRe classification and specify issues for assessing ASPs. METHODS: The retrospective study was conducted in a single center within an 845-bed hospital. Antimicrobial usage data for outpatients were obtained from medical records used for billing purposes. Antimicrobials for non-infectious treatment were defined by smaller dose of macrolides, tetracyclines with pemphigoid, rifaximin, and prophylactic sulfamethoxazole-trimethoprim (ST) agent. RESULTS: The usage of antimicrobials for non-infectious treatment increased from 25.3 % to 50.1 % for the ratio of the amount to defined daily doses (DDDs) and from 46.3 % to 65.9 % for prescription days between January 2015 and March 2024. The usage of prophylactic sulfamethoxazole-trimethoprim (ST) agents increased by 2.4 times, and the usage of rifaximin increased by more than 100 times. Macrolides for non-infectious treatment was stable or fluctuated while that for infection treatment decreased to that amount for non-infectious treatment. The ratios for Access increased from 31.9 % to 58 % and 42 % to 78 % by excluding the antimicrobials for non-infectious treatment. CONCLUSIONS: The findings suggested that the AWaRe classification might not be appropriate for assessing ASPs and comparing them among hospitals.

Digital Epidemiology of Prescription Drug References on X (Formerly Twitter): Neural Network Topic Modeling and Sentiment Analysis.

BACKGROUND: Data from the social media platform X (formerly Twitter) can provide insights into the types of language that are used when discussing drug use. In past research using latent Dirichlet allocation (LDA), we found that tweets containing "street names" of prescription drugs were difficult to classify due to the similarity to other colloquialisms and lack of clarity over how the terms were used. Conversely, "brand name" references were more amenable to machine-driven categorization. OBJECTIVE: This study sought to use next-generation techniques (beyond LDA) from natural language processing to reprocess X data and automatically cluster groups of tweets into topics to differentiate between street- and brand-name data sets. We also aimed to analyze the differences in emotional valence between the 2 data sets to study the relationship between engagement on social media and sentiment. METHODS: We used the Twitter application programming interface to collect tweets that contained the street and brand name of a prescription drug within the tweet. Using BERTopic in combination with Uniform Manifold Approximation and Projection and k-means, we generated topics for the street-name corpus (n=170,618) and brand-name corpus (n=245,145). Valence Aware Dictionary and Sentiment Reasoner (VADER) scores were used to classify whether tweets within the topics had positive, negative, or neutral sentiments. Two different logistic regression classifiers were used to predict the sentiment label within each corpus. The first model used a tweet's engagement metrics and topic ID to predict the label, while the second model used those features in addition to the top 5000 tweets with the largest term-frequency-inverse document frequency score. RESULTS: Using BERTopic, we identified 40 topics for the street-name data set and 5 topics for the brand-name data set, which we generalized into 8 and 5 topics of discussion, respectively. Four of the general themes of discussion in the brand-name corpus referenced drug use, while 2 themes of discussion in the street-name corpus referenced drug use. From the VADER scores, we found that both corpora were inclined toward positive sentiment. Adding the vectorized tweet text increased the accuracy of our models by around 40% compared with the models that did not incorporate the tweet text in both corpora. CONCLUSIONS: BERTopic was able to classify tweets well. As with LDA, the discussion using brand names was more similar between tweets than the discussion using street names. VADER scores could only be logically applied to the brand-name corpus because of the high prevalence of non-drug-related topics in the street-name data. Brand-name tweets either discussed drugs positively or negatively, with few posts having a neutral emotionality. From our machine learning models, engagement alone was not enough to predict the sentiment label; the added context from the tweets was needed to understand the emotionality of a tweet.