Text Mining for Translational Bioinformatics

Text mining tools are now widely used for the efficient management of information and resources in business, academic and research organizations. This paper provides a comprehensive overview of research articles on the application of text mining techniques in the field of Supply Chain Risk Management and the apparel industry. Research articles published between 2000 and 2020, were obtained from various journals through two online databases, i.e., SCOPUS and IEEE Xplore. Through a systematic approach following PRISMA guidelines, 370 research papers were screened, filtered and finally classified into three main areas: Supply Chain Risk Management and outsourcing in the apparel industry, application of text mining in Supply Chain Risk Management and application of text mining in the apparel industry. In this study, we have identified a comprehensive list of various available data sources for text mining, methodologies and risks associated with outsourcing in the apparel industry. We classify the gaps in expanding the application of text mining in the apparel industry’s Supply Chain Risk Management. Extracting useful information from online newspapers through text mining could vividly enhance the ability to monitor supply chain risks and provide the ability to link data to provide decision makers with the right information at the right time.

Large Language Models (LLMs) hold significant promise for improving clinical decision support and reducing physician burnout by synthesizing complex, longitudinal cancer Electronic Health Records (EHRs). However, their implementation in this critical field faces three primary challenges: the inability to effectively process the extensive length and fragmented nature of patient records for accurate temporal analysis; a heightened risk of clinical hallucination, as conventional grounding techniques such as Retrieval-Augmented Generation (RAG) do not adequately incorporate process-oriented clinical guidelines; and unreliable evaluation metrics that hinder the validation of AI systems in oncology. To address these issues, we propose CliCARE, a framework for Grounding Large Language Models in Clinical Guidelines for Decision Support over Longitudinal Cancer Electronic Health Records. The framework operates by transforming unstructured, longitudinal EHRs into patient-specific Temporal Knowledge Graphs (TKGs) to capture long-range dependencies, and then grounding the decision support process by aligning these real-world patient trajectories with a normative guideline knowledge graph. This approach provides oncologists with evidence-grounded decision support by generating a high-fidelity clinical summary and an actionable recommendation. We validated our framework using large-scale, longitudinal data from a private Chinese cancer dataset and the public English MIMIC-IV dataset. In these settings, CliCARE significantly outperforms baselines, including leading long-context LLMs and Knowledge Graph-enhanced RAG methods. The clinical validity of our results is supported by a robust evaluation protocol, which demonstrates a high correlation with assessments made by oncologists.

BackgroundErectile dysfunction is no longer a whisper in the shadows; it’s a rising tide threatening the sexual health of millions of men in different regions. In the cases of diabetes, the condition worsens and has a potent cocktail of physical and psychological distress, chipping away at men’s confidence, self-esteem, and mental health. This worrying trend shows no signs of slowing down, with projections claiming a staggering 322 million men globally could be affected in the near future. This urgent issue demands immediate attention and action. Thus, this umbrella review intended to estimate the current burden of erectile dysfunction and associated risk factors among diabetic patients in the global context.MethodsFollowing PRISMA guidelines, we searched for relevant studies in PubMed, Embase, Scopus, Web of Science, Cochrane Database of Systematic Reviews, and Google Scholar. These studies reported the prevalence of erectile dysfunction and associated risk factors in diabetic patients. The quality of the included studies was assessed using the Assessment of Multiple Systematic Reviews 2 tool. To estimate the pooled prevalence of erectile dysfunction, we employed a weighted inverse variance random-effects model. We further conducted subgroup analyses, assessed heterogeneity and publication bias, and performed sensitivity analyses to strengthen the robustness of our findings. Prediction intervals were also calculated to estimate the range within which future observations will likely fall. In all statistical analyses, the statistical significance was declared at P- value < 0.05.ResultsIn this umbrella review, a total of 108 030 male diabetic patients were included to estimate the global prevalence of erectile dysfunction. The pooled global prevalence of erectile dysfunction in diabetic patients was 65.8% (95CI: 58.3 − 73.3%), while In Africa it was 62.9% (95CI: 46.1–79.7). Age (> 40 years) (AOR = 1.95, 95CI: 1.03–3.24), DM duration (> 10years) (AOR = 1.90,95CI: 1.16–2.65), peripheral vascular disease (AOR = 2.74, 95CI: 1.42–4.06) and BMI (> 30 kg/m2) (AOR = 1.07,95CI: 1.01–1.20) were identified as associated risk factors of erectile dysfunction in diabetic patient.ConclusionThe high global prevalence of erectile dysfunction (ED) in diabetic patients is alarming, with an estimated two-thirds experiencing the condition. These findings underscore the significant burden of ED faced by diabetic men and emphasize the urgent need for global attention to the issue. This includes promoting early screening for erectile dysfunction in this population and ensuring access to appropriate treatment and support.

Electronic Health Records: Promises and Realities: A 3-Part Series Part I: The Digital Sea Change, Ready or Not

The key question of precision medicine is whether it is possible to find clinically actionable granularity in diagnosing disease and classifying patient risk. The advent of next-generation sequencing and the widespread adoption of electronic health records (EHRs) have provided clinicians and researchers a wealth of data and made possible the precise characterization of individual patient genotypes and phenotypes. Unstructured text-found in biomedical publications and clinical notes-is an important component of genotype and phenotype knowledge. Publications in the biomedical literature provide essential information for interpreting genetic data. Likewise, clinical notes contain the richest source of phenotype information in EHRs. Text mining can render these texts computationally accessible and support information extraction and hypothesis generation. This chapter reviews the mechanics of text mining in precision medicine and discusses several specific use cases, including database curation for personalized cancer medicine, patient outcome prediction from EHR-derived cohorts, and pharmacogenomic research. Taken as a whole, these use cases demonstrate how text mining enables effective utilization of existing knowledge sources and thus promotes increased value for patients and healthcare systems. Text mining is an indispensable tool for translating genotype-phenotype data into effective clinical care that will undoubtedly play an important role in the eventual realization of precision medicine.

This study aims to evaluate scientific publications on "Medical Informatics" and "Data Privacy" using a bibliometric approach to identify research trends, the most studied topics, and the countries and institutions with the highest publication output. The search was carried out utilizing the WoS Clarivate Analytics tool across SCIE journals. Subsequently, text mining, keyword clustering, and data visualization were applied through the use of VOSviewer and Tableau Desktop software. Between 1975 and 2023, a total of 7,165 articles were published on the topic of data privacy. The number of articles has been increasing each year. The text mining and clustering analysis identified eight main clusters in the literature: (1) Mobile Health/Telemedicine/IOT, (2) Security/Encryption/Authentication, (3) Big Data/AI/Data Science, (4) Anonymization/Digital Phenotyping, (5) Genomics/Biobank, (6) Ethics, (7) Legal Issues, (8) Cloud Computing. On a country basis, the United States was identified as the most active country in this field, producing the most publications and receiving the highest number of citations. China, the United Kingdom, Canada, and Australia also emerged as significant countries. Among these clusters, "Mobile Health/Telemedicine/IOT," "Security/Encryption/Authentication," and "Cloud Computing" technologies stood out as the most prominent and extensively studied topics in the intersection of medical informatics and data privacy.

The publication date is one day earlier then the EST date to provide the proceedings to attendees in Australian on the first day of the conference It is our great pleasure to welcome you to the ACM Ninth International Workshop on Data and Text Mining in Biomedical Informatics (DTMBIO'15), in conjunction with ACM Conference of Information and Knowledge Management (CIKM'15). As biomedical "big-data" are rapidly growing in quantity and quality, computational techniques to handle, store and analyze such information to acquire more valuable knowledge is now essential for relevant researchers. Along with the traditional medical records including electronic medical records (EMR), clinical test results and environmental records, high-throughput observations upon biological entities such as genes, transcripts, proteins and metabolites are coupled with individual phenotype information to infer elements that are associated with a specific phenomenon of our interest. This also requires an exploration of previously reported knowledge that exists in a form of canonical database and literature, to build a hypothesis and confirm the consistency. Accordingly, it is doubtless that application of cutting-edge techniques in data and text mining is a key for efficient biomedical knowledge generation in this "big-data" era. Since its inaugural meeting in 2006, started as the Text-mining in bioinformatics (TMBIO), DTMBIO has been continuously bringing together multidisciplinary pioneers from computer and biomedical communities. Over 100 high-quality studies have been presented in the past DTMBIO workshops, many of which are finally published as full paper in scientific journals. Although highly diverse in terms of methods, information sources and target phenotypes (diseases), we have been maintaining two major themes: (1) method development and improvement for data and text mining in bioinformatics that forms a basis for more efficient and accurate analyses, and (2) novel biomedical discoveries from high-throughput data based upon data and text mining techniques. We regard the balanced combination between pure computer science and biology as the primary factor for being successful in holding a creative and innovative workshop for both groups. And we anticipate this tradition will be kept for upcoming DTMBIO workshops too. The papers accepted for presentation and publication in this volume cover a variety of topics, including bio-text mining, bio-data mining, pathway and bio-network analysis, bio-ontology management, sequence analysis for massively parallel sequencing, protein-protein interactions, biomedical data classification, and biomedical information retrieval. We hope that these proceedings will serve as a valuable and up-to-date reference about the application of data- and text-mining techniques within biomedical informatics.

This paper reviews the challenges and opportunities associated with implementing health informatics in rural healthcare settings. It examines the unique barriers rural communities face, such as infrastructure limitations, workforce shortages, and financial constraints, which can impede the adoption and utilization of health informatics solutions. However, it also explores the promising opportunities afforded by health informatics, including telemedicine, electronic health records (EHRs), and mobile health technologies, to overcome these challenges and improve healthcare access, quality, and efficiency in rural areas. By addressing these challenges and leveraging the opportunities presented by health informatics, stakeholders can work towards achieving equitable healthcare outcomes for rural populations. By addressing these challenges and leveraging the opportunities presented by health informatics, stakeholders can work towards achieving equitable healthcare outcomes for rural populations. Health informatics holds immense potential to revolutionize healthcare delivery, particularly in rural settings where access to quality care is often limited Health informatics holds immense potential to revolutionize healthcare delivery, particularly in rural settings where access to quality care is often limited. Health informatics holds immense potential to revolutionize healthcare delivery, particularly in rural settings where access to quality care is often limited. Telemedicine, for instance, enables remote consultations and diagnostic services, breaking down geographic barriers and providing rural patients with access to specialists and healthcare expertise they may not have locally. Electronic Health Records (EHRs) streamline administrative processes, improve care coordination, and enhance patient safety by ensuring that vital health information is readily accessible to healthcare providers regardless of location. Mobile health technologies, including smartp hone apps and wearable devices, empower rural residents to monitor their health, track chronic conditions, and engage in self-management activities from the comfort of their homes.By embracing these innovations and overcoming infrastructure challenges, stakeholders in rural healthcare can significantly improve healthcare access, quality, and efficiency. Additionally, leveraging health informatics can help address workforce shortages by enabling healthcare providers to deliver more effective and integrated care, even in remote areas. Ultimately, by harnessing the opportunities presented by health informatics while addressing the unique challenges of rural healthcare, stakeholders can work towards achieving equitable healthcare outcomes for rural populations. Keywords: Health Informatics, Rural Healthcare, Electronic Health Records (EHRS), Mobile Health, Challenges.

BackgroundAdverse events induced by drug-drug interactions are a major concern in the United States. Current research is moving toward using electronic health record (EHR) data, including for adverse drug events discovery. One of the first steps in EHR-based studies is to define a phenotype for establishing a cohort of patients. However, phenotype definitions are not readily available for all phenotypes. One of the first steps of developing automated text mining tools is building a corpus. Therefore, this study aimed to develop annotation guidelines and a gold standard corpus to facilitate building future automated approaches for mining phenotype definitions contained in the literature. Furthermore, our aim is to improve the understanding of how these published phenotype definitions are presented in the literature and how we annotate them for future text mining tasks.ResultsTwo annotators manually annotated the corpus on a sentence-level for the presence of evidence for phenotype definitions. Three major categories (inclusion, intermediate, and exclusion) with a total of ten dimensions were proposed characterizing major contextual patterns and cues for presenting phenotype definitions in published literature. The developed annotation guidelines were used to annotate the corpus that contained 3971 sentences: 1923 out of 3971 (48.4%) for the inclusion category, 1851 out of 3971 (46.6%) for the intermediate category, and 2273 out of 3971 (57.2%) for exclusion category. The highest number of annotated sentences was 1449 out of 3971 (36.5%) for the “Biomedical & Procedure” dimension. The lowest number of annotated sentences was 49 out of 3971 (1.2%) for “The use of NLP”. The overall percent inter-annotator agreement was 97.8%. Percent and Kappa statistics also showed high inter-annotator agreement across all dimensions.ConclusionsThe corpus and annotation guidelines can serve as a foundational informatics approach for annotating and mining phenotype definitions in literature, and can be used later for text mining applications.

Background The electronic medical record (EMR) offers unique possibilities for clinical research, but some important patient attributes are not readily available due to its unstructured properties. We applied text mining using machine learning to enable automatic classification of unstructured information on smoking status from Swedish EMR data. Methods Data on patients’ smoking status from EMRs were used to develop 32 different predictive models that were trained using Weka, changing sentence frequency, classifier type, tokenization, and attribute selection in a database of 85,000 classified sentences. The models were evaluated using F-score and accuracy based on out-of-sample test data including 8500 sentences. The error weight matrix was used to select the best model, assigning a weight to each type of misclassification and applying it to the model confusion matrices. The best performing model was then compared to a rule-based method. Results The best performing model was based on the Support Vector Machine (SVM) Sequential Minimal Optimization (SMO) classifier using a combination of unigrams and bigrams as tokens. Sentence frequency and attributes selection did not improve model performance. SMO achieved 98.14% accuracy and 0.981 F-score versus 79.32% and 0.756 for the rule-based model. Conclusion A model using machine-learning algorithms to automatically classify patients’ smoking status was successfully developed. Such algorithms may enable automatic assessment of smoking status and other unstructured data directly from EMRs without manual classification of complete case notes.

Hematology Patient Explorer - an Informatics Infrastructure and Web Application Empowering Clinicians and Researchers to Use Electronic Health Record Information to Identify Clinical Cohorts of Interest

Objectives: Postprandial hyperglycaemia has emerged as widely accepted cardiovascular risk factor in diabetic patients. Acceptable postprandial glycemic control is easier to achieve by intensified insulin therapie by using insulin analogues than by using normal human insulin. Patients benefit from the possibility of injecting insulin analogues after meals. There was controversy, however, whether insulin resistance in type 2 diabetic patients might hamper postprandial glycemic control by insulin analogues.

BackgroundLuseogliflozin, a sodium–glucose cotransporter 2 inhibitor, potentially exerts pleiotropic effects on the liver. However, the sufficient evidence is still lacking. We aimed to investigate the effects of luseogliflozin on hepatic steatosis, fibrosis, and cardiometabolic risk factors in diabetic patients by a pooled meta-analysis.MethodsIn this pooled meta-analysis, we enrolled diabetic patients who participated in phase III clinical trials of luseogliflozin (luseogliflozin group n = 302, placebo group n = 191). The primary outcomes were changes in fatty liver index (FLI) and Hepamet fibrosis score (HFS) after 24 weeks. The secondary outcomes were changes in cardiometabolic risk factors after 24 weeks. Statistical analysis was performed using propensity scoring analysis by the inverse probability of treatment weighting method.ResultsPrimary outcomes: Luseogliflozin significantly decreased FLI compared to placebo after 24 weeks (adjusted coefficient − 5.423, 95%CI − 8.760 to − 2.086, P = 0.0016). There was no significant difference in changes in HFS between the two groups. However, luseogliflozin significantly decreased HFS compared to placebo in diabetic patients with ALT > 30 U/L (adjusted coefficient − 0.039, 95%CI − 0.077 to − 0.001, P = 0.0438) and with FIB-4 index > 1.3 (adjusted coefficient − 0.0453, 95%CI − 0.075 to − 0.016, P = 0.0026). Secondary outcom8es: Luseogliflozin significantly decreased HbA1c level, HOMA-IR value, BMI, and uric acids level, and increased HDL cholesterol level compared to placebo.ConclusionsThis pooled meta-analysis demonstrated that 24-week treatment with luseogliflozin improved hepatic steatosis and fibrosis indexes in diabetic patients, especially those with liver injury. Furthermore, luseogliflozin improved various cardiometabolic risk factors. Thus, luseogliflozin may be useful for improving MASLD in diabetic patients.Graphical abstract

Identification and prioritisation of diabetic nephropathy risk factors in diabetes patients using machine learning approach

Cardiovascular disease (CVD) is the major cause of morbidity and mortality in diabetes. To determine the proportion of patients who met the American Diabetes Association guidelines for control of CVD risk factors and to assess the achievement of these guidelines in women compared with men, we conducted a cross-sectional study of 3678 diabetic cohorts followed at seven medical centers, two Veteran Administration hospitals, three urban clinics, and two suburban clinics. Overall, 28% met the target blood pressure of <130/80 mm Hg, 48.8% achieved a goal low-density lipoprotein cholesterol of <100 mg/dL, and 35.8% had hemoglobin A1c of <7%. Gender comparisons of 2788 diabetic patients at urban and suburban centers showed that women had a lower percentage of low-density lipoprotein cholesterol <100 mg/dL (45.8 vs. 51.3, p<0.01) and a lower percentage of screening for retinopathy (54 vs. 60, p<0.01) and nephropathy (37 vs. 49, p<0.01). However, overall there were no gender differences in the percentage of patients who achieved a goal blood pressure <130/80 mm Hg or hemoglobin A1c <7%. Control of blood pressure and other CVD risk factors in diabetic patients was largely suboptimal, especially for diabetic women. These observations underscore the need for better strategies for control of CVD risk in the diabetic population in general, and women in particular.