Scholarlink

Network analysis provides a complementary lens to classical diagnostics by revealing co-occurrence patterns among clinical, laboratory, demographic, vaccination, and comorbidity features that define disease profiles. Using 324 notified meningitis cases from the Sinan dataset (2003–2022), we built etiology-specific networks for meningococcal meningitis, tuberculous meningitis, Haemophilus influenzae meningitis. Etiology-specific anchors confined to one disease included Neisseria meningitidis isolation in CSF culture (meningococcal), fever and Haemophilus influenzae detection in CSF culture or antigen testing (H. influenzae), and adult age group (tuberculous). Bridging nodes such as vomiting, cloudy CSF, and pediatric age group (shared between H. influenzae and meningococcal) linked otherwise distinct networks, revealing points of diagnostic overlap. Demographic patterns showed pediatric predominance in H. influenzae and meningococcal meningitis, and adult predominance in tuberculous meningitis, alongside sex-linked trends. Vaccination nodes reflected both protection and residual disease in under-immunized groups, while HIV/AIDS and tuberculosis anchored the tuberculous network. These relational patterns connect structural network metrics with clinical meaning, offering interpretable insights to refine differential diagnosis and inform graph-based decision-support tools.

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Co-Occurrence Network Analysis of Clinical Features in Meningitis: Insights into Symptom Clustering and Etiological Patterns

  • 会议论文

  • 发表日期：

    2026年

查看全文
  • A. Messai,
  • A. Drif,
  • A. Ouyahia,
  • M. Rais,
  • M. Guechi,
  • L. Kaderali,
  • H. Cherifi

摘要

Network analysis provides a complementary lens to classical diagnostics by revealing co-occurrence patterns among clinical, laboratory, demographic, vaccination, and comorbidity features that define disease profiles. Using 324 notified meningitis cases from the Sinan dataset (2003–2022), we built etiology-specific networks for meningococcal meningitis, tuberculous meningitis, Haemophilus influenzae meningitis. Etiology-specific anchors confined to one disease included Neisseria meningitidis isolation in CSF culture (meningococcal), fever and Haemophilus influenzae detection in CSF culture or antigen testing (H. influenzae), and adult age group (tuberculous). Bridging nodes such as vomiting, cloudy CSF, and pediatric age group (shared between H. influenzae and meningococcal) linked otherwise distinct networks, revealing points of diagnostic overlap. Demographic patterns showed pediatric predominance in H. influenzae and meningococcal meningitis, and adult predominance in tuberculous meningitis, alongside sex-linked trends. Vaccination nodes reflected both protection and residual disease in under-immunized groups, while HIV/AIDS and tuberculosis anchored the tuberculous network. These relational patterns connect structural network metrics with clinical meaning, offering interpretable insights to refine differential diagnosis and inform graph-based decision-support tools.