Informatics in Medicine Unlocked最新文献

While time series are extensively utilized in economics, finance and meteorology, their application in epidemics has been comparatively limited. To facilitate a comprehensive research endeavor on this matter, we deemed it necessary to commence with a systematic literature review (SLR). This Systematic Literature Review aims to assess, based on a sample of relevant papers, the use of Time Series Methods (TSM) in epidemic prediction, with a special focus on African issues and the impact of COVID-19. The SLR was conducted using databases such as ACM, IEEE, PubMed and Science Direct. Open access published papers in English, in a pear reviewed Journals, from 2014 to 2023, containing keywords such as Time Series, Epidemic and Prediction were selected. The findings were summarized in an adapted PRISMA flow diagram. We end up with a sample of 36 papers. As conclusion, TSM are not so used in epidemic prediction as in some other domains, even though epidemic data are collected as time series. Just very few works address African issues regarding diseases and countries. COVID-19 is the pandemic that revealed and enhanced the used of TSM to forecast epidemics. This work paves ways for R&D on epidemiology, based on TSM.

Hormone peptides are small signaling molecules that regulate key cellular processes such as cell growth, and differentiation. Hormone peptide identification is important for understanding their potential associations with certain diseases such as attention deficit hyperactivity disorder, diabetes, and psychiatric disorders. A comprehensive understanding of hormone peptides’ roles in cellular signaling, and immune regulation can provide insights into their therapeutic potential. Hormone peptides are identified through wet-lab approaches which are restricted by resource-intensive processes, limited scalability, and cost ineffectiveness. In an effort to substitute experimental approaches with computational predictors, researchers leveraged the capabilities of machine learning (ML) classifiers. These classifiers have inherent dependency over statistical vectors that are generated by extracting amino acids’ distinctive patterns from peptide sequences. Classifiers utilize these vectors for discriminating peptides into hormone and non-hormone classes. However, the performance of current predictors is constrained due to their inability to effectively extract discriminative amino acids patterns from peptide sequences. Following the need for a powerful predictor, the paper in hand presents a novel sequence encoder namely, CTD-G that transforms peptide sequences into statistical vectors by extracting 3 different types of amino acids patterns namely composition, transition, and distribution. Across public benchmark dataset, the proposed CTD-G encoder potential is compared with 56 existing encoders under two different evaluation strategies namely intrinsic and extrinsic. In Intrinsic evaluation, TSNE-based visualization demonstrates reduced overlap between clusters of hormone and non-hormone peptides with the proposed encoder’s statistical vectors compared to existing encoders. Extrinsic evaluation demonstrates the superiority of the proposed encoder, as 7 out of 11 ML classifiers achieve better performance with its statistical vectors compared to those from existing encoders. Furthermore, the proposed predictor outperforms existing hormone peptide classification predictors by 1.5% in accuracy, 5.36% in sensitivity, 1.80% in specificity, and 2.62% in MCC. To facilitate the scientific community, a web application is available at https://sds_genetic_analysis.opendfki.de/.

A total of 116 metabolites of Caulerpa racemosa were identified. However, only three (DHA, EPA, and EPAS) were found to have high anti-inflammatory potential, with Pa scores ranging from 0.764 to 0,827. The inhibition constant (Ki) and binding energy interactions with COX-2 revealed by DHA (−8.83 kcal/mol: 0.338 μM), EPA (−8.35 kcal/mol: 0.763 μM), EPAS (−8.05 kcal/mol: 1.25 μM). They were used to bind to the fundamental residues of COX-2 (TYR 348, VAL 349, LEU 384, TYR 385, and TRP 387). The result of molecular dynamics showed that DHA, EPA, and EPAS had high stability while interacting with COX-2 in 310 K. The stabilities were 1.8 Å for DHA from 60 Ns to 200 Ns, 2.0 Å for EPA from 75 Ns to 200 Ns, and 2.2 Å for EPAS from 100 Ns to 200 Ns. Additionally, the potential energy of DHA (−1.069.250 eV) was higher compared with that of EPA (−1.069.247 eV) and EPAS (−1.069.220 eV). This data shows that DHA, EPA, and EPAS could stably inhibit COX-2 by blocking the transcriptional regulation of COX-2 via TYR348, VAL349, LEU384, TYR385, and TRP387.

Early detection of cardiovascular diseases (CVDs) is crucial for managing cardiovascular diseases and improving patient outcomes. Deep neural networks have the potential to reduce the reliance on costly and time-consuming clinical tests, leading to cost savings for patients and healthcare systems. This study proposes the development of specialized convolutional neural networks for the automated selection of essential variables, employing various preprocessing procedures. It evaluates the approach using the UCI repository heart disease dataset, focusing on early-stage heart disease identification to enhance early prediction and intervention for CVD. To address the challenge of achieving higher accuracy, we introduce an approach using one-dimensional convolutional neural networks, incorporating extensive testing to optimize the network architecture and enhance predictive performance. Additionally, recognizing the impact of features on accuracy, a comprehensive data analysis was performed. Through a meticulous selection process, we identified and utilized key features that significantly influenced the accuracy of our model, contributing to more reliable predictions. Finally, cross-validation techniques were implemented to precisely evaluate the efficacy of our work. Numerous experiments were conducted to demonstrate the relevance of our research. The prediction accuracy was found to be 99.95% when employing a train–test approach, while it was approximately 98.53% when employing K-Fold cross-validation. In comparison to existing literature, our approach outperforms a recent best study that proposed a Catboost model, achieving an F1-score of about 92.3% and an average accuracy of 90.94%. This signifies a substantial improvement in predictive performance, with a percentage improvement of approximately 9.90% compared to the Catboost model.

The recent coronavirus pandemic impacted the mental health of people worldwide as it rapidly spread to most countries. Social media has provided a cost-effective method to examine these negative effects, but which aspects of the disease impacted the general population's psychology is not well understood. This study examined one potential factor that moderated people's responses to the fast-spreading, deadly coronavirus disease (COVID) during its emergence from January 2020 to March 2020. Applying sentiment analysis to 3.2 million COVID-related messages posted on Twitter from 189 countries, we examined how the physical distance to COVID impacted the attention and emotions of the general population as it spread around the globe. The spatial distance from each message's origin country to the nearest COVID-infected country was computed to use as an independent variable. Statistical analyses revealed that spatial distance significantly influenced both public attention and sentiment toward COVID, even when controlling for confounders. As the disease came closer, more tweets were posted and the average sentiment became more negative. These observations suggest that physical proximity to a threat influences how much attention people pay to the threat and how they respond to it emotionally. This is in line with previous disaster research and fits the psychological framework of construal level theory. Although these findings are limited in their generalizability, they have important implications. In practice, communicating the personal risks of a disease outbreak to distant people might increase public engagement in protective behaviors such as social distancing and hand washing, subsequently slowing disease spread.

Artificial Intelligence (AI) refers to the development of computer systems that can perform tasks that typically require human intelligence. The utilization of AI in healthcare, particularly in dental clinics, has drawn attention to the issue of appointment no-shows. These no-shows have detrimental effects such as increased waiting times, limited-service access, and financial burden on healthcare providers. Therefore, optimizing the organization of dental clinics is crucial to effectively cater to a diverse patient population with varying dental needs, especially considering the projected rise in demand for dental care. To address the problem of appointment no-shows, the researchers proposed a programming model that harnesses machine learning algorithms. Three specific algorithms, namely Decision Trees, Random Forest, and Multilayer Perceptron, were employed, with the Multilayer Perceptron being used for the first time in this particular context. The researchers collected a dataset from five dental facilities specializing in nine areas and employed Explainable AI techniques to gain insights into the factors contributing to patient absences. The model's performance was evaluated using multiple metrics. The Decision Tree model exhibited favorable accuracy, achieving 79% precision, 94% recall, 86% F1-Score, and 84% AUC (Area Under the Curve). The Random Forest model demonstrated even higher accuracy, with 81% precision, 93% recall, 87% F1-Score, and 83% AUC. Similarly, the Multilayer Perceptron model attained an accuracy of 80% precision, 91% recall, 86% F1-Score, and 83% AUC.

The investigation of hydroxyxanthone derivatives has been conducted, including molecular docking, molecular dynamics simulation MM-PBSA binding energy calculation, and pharmacokinetics prediction of the potential plasmodium falciparum dihydrofolate reductase (pfDHFR) and plasmodium falciparum dihydroorotate dehydrogenase (pfDHODH) inhibitor. The Docking result showed that compound 1,3,6,7-tetrahydroxy-5,8-bis(3-methyl-2-buten-1-yl)-9H-xanthen-9-one (X16) was found to be the best ligand with good inhibitory action against pfDHFR. Meanwhile, the pfDHODH protein was compounded 1,3,6,7-tetrahydroxy-5,8-dinitro-9H-xanthen-9-one (X14). Additionally, the hydroxyxanthone X16 complex showed more excellent stability in the molecular dynamics simulation of the pfDHFR protein than the ligand WR99210 and chloroquine. The MM-PBSA calculation showed that compound X16 had lower binding energy than ligand WR99210. However, 1,3-dihydroxy-8-(3-methyl-2-buten-1-yl)-9H-xanthen-9-one (X4), 1,3,6,7-tetrahydroxy-8-nitro-9H-xanthen-9-one (X10), 1,3,6,7-tetrahydroxy-9-oxo-9H-xanthene-8-sulfonic acid (X11), and 1,3,6,7-tetrahydroxy-5,8-dinitro-9H-xanthen-9-one (X14) complexes were shown to be more stable than chloroquine and to have the same stability when compared to the native ligand A26, according to a molecular dynamics simulation conducted in pfDHODH protein. The MM-PBSA calculation showed that compound X14 had lower binding energy than ligand A26. The hydroxyxanthones X4, X10–11, X14, and X16 fulfill Lipinski's rule parameters in terms of physicochemical and ADMET qualities and parameters related to absorption, distribution, metabolism, excretion, and toxicity tests. To sum up, hydroxyxanthones X4, X10–11, X14, and X16 have the potential to be antimalarial medications, but more in vivo and in vitro testing is needed to confirm this.

Ebola, one of the deadliest known infectious diseases, was the root of epidemics in Western Africa from 2013 to 2016. Like other deadly viruses in the family Filoviridae with a high fatality rate, this virus also causes hemorrhagic fever. As a result, the Ebola virus (EBOV) represents a threat to global health. Since there are currently no effective treatments for EBOV infections, this study aims to identify potential natural drug candidates that may block the EBOV VP40 to prevent Ebola infections. The compounds were analyzed using ADMET, molecular docking, post-docking MM-GBSA, and molecular dynamics (MD) simulations. ADMET analysis identified 187 out of 452 compounds. According to molecular docking, the best three compounds were chosen from 187 compounds for further study with binding affinity −8.469, −8.175, and −7.918 kcal/mol for CID_21721878 (Kushenol L), CID_133561472 (2-[2,4-dihydroxy-5-[2-(2-hydroxypropan-2-yl)-5-methylphenyl]phenyl]-5,7-dihydroxy-2,3-dihydrochromen-4-one), and Amb_29844215 (Cathayanon I), respectively. The lead three compounds coordinated with the protein's shared amino acid residues (ILE216, PRO286, VAL287, LEU288, LEU213, PRO146, and VAL100) during molecular docking with hydrophobic bonds. Then, molecular docking results were validated using post-docking MM-GBSA of those three compounds are Kushenol L, 2-[2,4-dihydroxy-5-[2-(2-hydroxypropan-2-yl)-5-methylphenyl]phenyl]-5,7-dihydroxy-2,3-dihydrochromen-4-one and Cathayanon I had negative binding free energies of −69.53, −52.85, and −59.74 kcal/mol, respectively. All the selected compounds exhibit favorable pharmacokinetic (Pk) and toxicological properties, supporting their safety and efficacy. These three compounds were further evaluated using MD simulation, confirming the compounds' binding stability to the desired protein. After MD simulation, PCA, and DCCM analysis were performed. From all of these can suggest the best compound which is CID_21721878 (Kushenol L), which is a phytochemical derived from Cannabis sativa, another one is CID_13356472 which comes after Kushenol L, which is also a phytochemical found in several plants: Maclura tricuspidate, Euchresta japonica, Maclura pomifera. Both compounds can potentially inhibit EBOV VP40 protein activity.

The complexity of the facilities of healthcare providers goes beyond their physical articulation, function, and organization; it also involves integrating technology and healthcare activities that continuously evolve due to medical research and technological advancements. As a result, hospitals require a flexible approach that can accommodate the changing demands of patients, medical professionals, and researchers. This flexibility is essential in ensuring that hospitals can meet the diverse needs of their users and adapt to fast-changing medical requirements. Therefore, integrating analytical capabilities of Machine Learning algorithms in healthcare services is a vital aspect of Flexible Healthcare Systems. Furthermore, it enables hospitals to efficiently organize patient data and optimize treatment plans by analyzing vast amounts of patient data. In this paper, we explored the role of Machine Learning by applying Deep Convolutional Neural Networks on three unique datasets to predict the risk of developing cancer using health informatics and to demonstrate how computer-based vision can improve cancer prognosis by analyzing medical images. Furthermore, we have employed advanced CNNs for high-accuracy cancer detection in images, using a streamlined model that combines feature-detecting convolutional layers with complexity-reducing pooling layers which ensures effective cancer identification. The implementation of these models into healthcare delivery can potentially improve patient outcomes and system-level efficiencies, but carefully considering their limitations and ethical implications are essential.