• HOME
  • NEWS
  • EXPLORE
    • CAREER
      • Companies
      • Jobs
    • EVENTS
    • iGEM
      • News
      • Team
    • PHOTOS
    • VIDEO
    • WIKI
  • BLOG
  • COMMUNITY
    • FACEBOOK
    • INSTAGRAM
    • TWITTER
Tuesday, November 17, 2026
BIOENGINEER.ORG
No Result
View All Result
  • Login
  • HOME
  • NEWS
  • EXPLORE
    • CAREER
      • Companies
      • Jobs
        • Lecturer
        • PhD Studentship
        • Postdoc
        • Research Assistant
    • EVENTS
    • iGEM
      • News
      • Team
    • PHOTOS
    • VIDEO
    • WIKI
  • BLOG
  • COMMUNITY
    • FACEBOOK
    • INSTAGRAM
    • TWITTER
  • HOME
  • NEWS
  • EXPLORE
    • CAREER
      • Companies
      • Jobs
        • Lecturer
        • PhD Studentship
        • Postdoc
        • Research Assistant
    • EVENTS
    • iGEM
      • News
      • Team
    • PHOTOS
    • VIDEO
    • WIKI
  • BLOG
  • COMMUNITY
    • FACEBOOK
    • INSTAGRAM
    • TWITTER
No Result
View All Result
Bioengineer.org
No Result
View All Result
Home NEWS Science News Health

Bibliometric Study Maps Decade of Machine Learning Advances in Lupus Research

by
November 11, 2026
in Health
Reading Time: 7 mins read
0
Bibliometric Study Maps Decade of Machine Learning Advances in Lupus Research

Bibliometric Study Maps Decade of Machine Learning Advances in Lupus Research

Share on FacebookShare on TwitterShare on LinkedinShare on RedditShare on Telegram

A comprehensive bibliometric evaluation published in Egyptian Rheumatology and Rehabilitation reveals that scientific literature applying machine learning techniques to systemic lupus erythematosus expanded thirtyfold between 2015 and 2025. Researchers Datian Fu, Linyan Tan, and Qiang Yang conducted a visual and quantitative scientometric mapping to trace publication trajectories, collaborative linkages, foundational knowledge structures, and emerging conceptual frontiers across the past decade. Systemic lupus erythematosus represents a chronic, heterogeneous autoimmune disorder characterized by systemic inflammation, pathogenic autoantibody generation, and substantial potential for multi-organ damage. Investigators have turned to machine learning as an analytical tool capable of parsing high-dimensional biological and clinical data to enhance diagnostic accuracy, subtype heterogeneous patient groups, uncover biological pathways, and predict adverse outcomes. However, the accompanying surge in published literature has produced a fragmented and redundant scientific landscape, creating an urgent requirement for systematic synthesis to distinguish validated clinical innovations from preliminary computational exercises.

To rigorously reconstruct the development of this interdisciplinary field, the investigators retrieved relevant literature from the Web of Science Core Collection spanning from January 1, 2015, through December 17, 2025. The database search coupled comprehensive topic terms for systemic lupus erythematosus, systemic lupus, and disseminated lupus with algorithmic designations including artificial intelligence, machine learning, deep learning, convolutional neural networks, neural networks, random forests, support vector machines, and gradient boosting. The query additionally utilized the Cited Reference Search function to capture foundational, highly cited seminal papers that keyword filters might otherwise omit. An initial yield of 454 records underwent duplicate resolution and two-stage dual-independent screening, resolving discrepancies by consensus or consultation with a third senior researcher. Eligible records comprised English-language research and review articles, while meeting abstracts, letters, editorial materials, book chapters, early access items, corrections, and entries lacking full texts, keywords, or abstracts were excluded. The final dataset of 451 publications was examined using CiteSpace analytical software version 6.3.R2 alongside Microsoft Office Excel 2025 to evaluate annual output, geographic networks, institutional ties, authorship constellations, co-cited references, co-cited journals, and keyword citation bursts.

The quantitative trajectory of research output demonstrated continuous exponential growth across the examined eleven-year interval, rising from 4 publications in 2015 to 121 papers in 2025. The authors fitted an exponential regression model to the annual publication volumes, yielding the equation y = 5.7107e^(0.3447x) with a coefficient of determination of 0.8447. This strong mathematical fit quantitatively confirms an accelerating rate of inquiry, driven particularly by five influential papers published between 2019 and 2021 that catalyzed subsequent investigations. For temporal and network evaluation, CiteSpace divided the 2015 to 2025 timespan into annual slices and implemented the g-index with a scaling factor of k = 25 to balance the annual representation of highly cited and emerging articles against citation skewness. Influential network positions were detected using node betweenness centrality, where scores exceeding 0.1 identify pivotal coordinating nodes. Topological networks were refined using the Pathfinder and Pruning sliced networks algorithms, while network clusters were extracted using log-likelihood ratio algorithms within latent semantic indexing, and network normalization was performed using the Jaccard index and Cosine association strength.

Geographic and institutional analyses indicated that global productivity is predominantly anchored by the People’s Republic of China and the United States, which serve as central nodes in an international cooperation network containing 50 country nodes and 243 collaborative connections. China produced the greatest overall volume of publications, representing the only developing nation ranked among the top five contributor countries. The authors observed that this volume corresponds with substantial domestic disease burden, national research funding prioritization, and targeted strategic state investments in artificial intelligence and biomedical technologies. Conversely, the United States maintained central scientific leadership and established extensive transatlantic partnerships with researchers across the United Kingdom, Sweden, Switzerland, and France, supported by established biotechnology infrastructure and collaborative academia-industry initiatives. China developed distinct collaborative ties with investigators in Japan, Spain, and Iran. Among 249 institutions connected by 578 collaborative edges, the University of California System and Harvard University led global institutional output with 13 publications each, closely followed by Nanjing University with 12 contributions. Five of the top ten publishing institutions were located in the United States, three in China, and two in England.

The authorship network revealed Peter E. Lipsky and Amrie C. Grammer as the most prolific investigators in the domain, each contributing 9 publications to the literature corpus. However, despite their substantial individual contributions and high publication volume, CiteSpace analysis uncovered generally low betweenness centrality across the author network. This topological pattern demonstrates that collaborative links among independent research groups remain relatively dispersed and fragmented. The authors emphasized that the observed sparsity of cross-institutional authorship clusters highlights an ongoing structural need for higher-quality, large-scale international consortia. Establishing unified, multi-center research partnerships will be essential to assemble the harmonized, diverse patient cohorts required to train and validate machine learning architectures that avoid localized cohort biases and generalize successfully across distinct geographic and demographic populations.

Analysis of co-cited references mapped a structural knowledge base comprising 553 cited documents connected through 1,606 co-citation links, yielding a network density of 0.0105. The single most cited reference in the network was the 2019 publication by Aringer and colleagues, which accumulated 30 co-citations. This standardized diagnostic framework provided an objective foundation for downstream computational investigations. A second landmark co-cited study developed an automated diagnostic algorithm using least absolute shrinkage and selection operator logistic regression, utilizing 14 clinical and serological features to compute individualized risk probabilities with an overall diagnostic accuracy of 94.8 percent. A third highly cited study examined machine learning classifiers designed to predict disease activity by integrating multi-cohort gene expression datasets. That investigation established that models trained on co-expression gene modules achieved greater robustness, demonstrating approximately 70 percent accuracy across technically variable platforms, whereas a random forest model achieving 83 percent cross-validation accuracy suffered notable performance degradation when encountering technical batch heterogeneity.

The journal co-citation landscape, comprising 503 publication venues and 2,768 citation links, demonstrated a cohesive core of specialized rheumatology journals complemented by prominent interdisciplinary and high-impact medical outlets. Core disciplinary publications—specifically Arthritis Research & Therapy, Journal of Rheumatology, Rheumatology, Arthritis & Rheumatology, Annals of the Rheumatic Diseases, and Lupus—served as foundational structural hubs anchoring early and ongoing computational inquiries. Concurrently, high-impact multidisciplinary and immunological venues such as Nature Reviews Rheumatology, Nature Communications, Autoimmunity Reviews, and Frontiers in Immunology occupied prominent co-citation positions, reflecting the translation of computational methodologies into broader biological contexts. The open-access journal PLOS ONE further contributed to broad-scope methodological dissemination. Temporal network tracking revealed a substantial increase in warm-toned citation nodes between 2022 and 2025, capturing an active and accelerating phase of journal-level integration. This dense convergence across rheumatology, clinical immunology, and data science indicates that machine learning is no longer treated merely as a niche computational technique, but rather as an integral analytical methodology addressing diverse clinical questions spanning diagnostic stratification, pathogenetic dissection, and therapeutic outcome forecasting.

Keyword co-occurrence analysis established a conceptual network of 371 nodes and 1,577 links organized into 10 distinct thematic clusters, dominated by high-frequency terms including systemic lupus erythematosus, machine learning, classification, disease, and artificial intelligence. Applying Kleinberg’s burst detection algorithm with a default parameter of gamma = 1.0 identified the top 25 terms experiencing rapid surges in citation frequency. Prominent burst terms included disease activity, type I interferon, prediction, women, mri, and pathogenesis. These lexical shifts mirror an evolving research focus moving away from static diagnostic classification toward dynamic, multi-modal monitoring. Recent investigations increasingly synthesize high-dimensional clinical parameters with quantitative magnetic resonance imaging and type I interferon gene signatures to track disease activity over time. Similarly, prognostic modeling efforts have expanded to incorporate sex-specific biological variables and non-apoptotic cell death mechanisms, such as ferroptosis in systemic inflammation. Rather than relying solely on individual laboratory measurements, contemporary computational workflows utilize cross-modal data fusion to construct composite predictive biomarkers that reflect the multifaceted biological processes underlying clinical disease flares and severe complications such as lupus nephritis.

Synthesizing keyword clustering, co-citation links, and publication trajectories, the authors delineated three core thematic domains defining the field: disease classification and diagnosis, prediction of disease activity and outcomes, and pathogenesis and biomarker discovery. In diagnostic classification, supervised models such as random forests and regularized regression have demonstrated classification accuracies reaching 93.79 percent when distinguishing systemic lupus erythematosus from related autoimmune conditions. In outcome prediction, quantitative pharmacodynamic frameworks—such as the maximum effect model applied to sirolimus therapy—and risk stratification algorithms have shown utility in anticipating therapeutic plateaus and predicting adverse pregnancy events. In pathogenetic research, machine learning applied to multi-omics data has helped uncover novel cellular pathways and molecular drivers. However, across all three thematic areas, the authors identified pervasive translational barriers, including data fragmentation across siloed clinical registries, a reliance on single-center datasets, and the prevalence of opaque black-box algorithms that fail to provide transparent biological rationales. Many complex computational models deliver only marginal diagnostic gains over conventional, interpretable clinical indices, while high-throughput omics analyses frequently encounter the curse of dimensionality, generating statistically robust associations that remain biologically unverified and unable to separate causal disease drivers from secondary bystander phenomena.

To overcome these systemic bottlenecks, the authors outlined critical methodological priorities necessary to transition machine learning in systemic lupus erythematosus from exploratory academic demonstrations to reliable clinical decision-support systems. Future efforts must prioritize the creation of explainable, multimodal architectures that integrate longitudinal clinical observations with dynamic multi-omics layers, replacing static cross-sectional classifiers with adaptive temporal forecasting tools. Concurrently, computational researchers should embrace hypothesis-driven paradigms that leverage machine learning algorithms to elucidate biological mechanisms rather than relying exclusively on unguided predictive pattern recognition. To resolve generalizability deficits, prospective models require rigorous external validation within expansive, multi-center, and ethnically diverse patient cohorts before point-of-care implementation can occur. Finally, the authors emphasized the necessity of bridging the phenotype-to-model gap by developing subtype-specific classifiers that directly accommodate lupus heterogeneity. The study noted its own methodological constraints, including reliance on the Web of Science Core Collection, restriction to English-language publications, citation biases such as self-citation, and software-specific algorithmic parameters, while concluding that the field’s future clinical success will depend fundamentally on collaborative, interpretable, and externally validated research designs.

Subject of Research: Scientometric and visual analysis of machine learning applications in systemic lupus erythematosus research from 2015 to 2025

Article Title: Machine learning in systemic lupus erythematosus research: a bibliometric and visual analysis (2015–2025)

Article References: Fu, D., Tan, L., & Yang, Q. (2026). Machine learning in systemic lupus erythematosus research: a bibliometric and visual analysis (2015–2025). Egyptian Rheumatology and Rehabilitation, 53(1), Article 53. https://doi.org/10.1186/s43166-026-00425-w

Image Credits: AI Generated

DOI: 10.1186/s43166-026-00425-w

Keywords: systemic lupus erythematosus, machine learning, bibliometric analysis, CiteSpace, Web of Science Core Collection, disease classification, disease activity prediction, biomarker discovery, lupus nephritis, type I interferon, co-citation analysis, pathogenesis, random forest, support vector machines, multimodal data, prognostic modeling, scientific collaboration

News Source: Teresa Odom. (November 11, 2026). Bibliometric Study Maps Decade of Machine Learning Advances in Lupus Research. Scienmag.

Tags: Bibliometric analysisbiomarker discoveryCiteSpaceco-citation analysisdisease activity predictiondisease classificationlupus nephritisMachine Learningmultimodal datapathogenesisprognostic modelingRandom ForestScientific Collaborationsupport vector machinessystemic lupus erythematosustype I interferonWeb of Science Core Collection
Share12Tweet7Share2ShareShareShare1

Related Posts

Cardiovascular Disease Significantly Increases Economic and Humanistic Burden in Kidney Failure

Cardiovascular Disease Significantly Increases Economic and Humanistic Burden in Kidney Failure

November 17, 2026
Thai Hypertension Study Reveals Shifts in Cardiovascular Risk Factors, 2011–2018

Thai Hypertension Study Reveals Shifts in Cardiovascular Risk Factors, 2011–2018

November 17, 2026

Delayed Thymopoiesis Follows Thymus Transplantation and Infection in Athymic Infant

November 17, 2026

Anemic Surgical Patients Suffer Greater Microvascular Flow Impairment Under Fluid Therapy

November 17, 2026

POPULAR NEWS

  • Pharmacist Prescribing for Urinary Tract Infections Shows High Symptom Resolution

    Pharmacist Prescribing for Urinary Tract Infections Shows High Symptom Resolution

    29 shares
    Share 12 Tweet 7
  • Bioactive Adrenomedullin Rises During Hemorrhagic Shock and Stabilizes With Fluid Resuscitation

    29 shares
    Share 12 Tweet 7
  • Serious Gaming Platform Validated for Basic Laparoscopic Skills Training

    29 shares
    Share 12 Tweet 7
  • Anti-Aging Drugs Show Sex-Specific Interleukin-11 Effects in Mouse Abdominal Fat

    29 shares
    Share 12 Tweet 7

About

We bring you the latest biotechnology news from best research centers and universities around the world. Check our website.

Follow us

Recent News

Dendritic Cell Virus-Like Vesicles Repurpose Arc Gene for Cancer Vaccination

Cardiovascular Disease Significantly Increases Economic and Humanistic Burden in Kidney Failure

Optimized Nanostructured Lipid Carriers Boost Abiraterone Activity in Prostate Cancer Cells

Subscribe to Blog via Email

Enter your email address to subscribe to this blog and receive notifications of new posts by email.

Join 88 other subscribers
  • Contact Us

Bioengineer.org © Copyright 2023 All Rights Reserved.

Welcome Back!

Login to your account below

Forgotten Password?

Retrieve your password

Please enter your username or email address to reset your password.

Log In
No Result
View All Result
  • Homepages
    • Home Page 1
    • Home Page 2
  • News
  • National
  • Business
  • Health
  • Lifestyle
  • Science

Bioengineer.org © Copyright 2023 All Rights Reserved.