Scholarlink

<p>Reference intervals (RIs) are the universal established methodology for interpreting numerical clinical data by comparing individual test results to population-based benchmarks. The quality of these RIs significantly influences individual-level decision-making. This study aims to answer the question on how we can effectively leverage an individual’s personal data in conjunction with that of their peers to compute individual reference intervals (IRIs) for key clinical parameters relevant to disease detection and progression. We describe the IRIS workflow that includes prior data processing and data quality check procedures. The computation of IRI involves the test results of multiple “healthy” data points from the same subject(s) and also from the peers. The model adjusts for covariates like sex and age, enhancing accuracy. The workflow demonstrated the potential utility of IRI in clinical and omics data from two longitudinal studies. For healthy populations, IRIs showed diagnostic value in chronic diseases, while in diseased cohorts, they enabled effective disease monitoring. An integrated application IRIS has been developed, incorporating all described steps in an easy-to-use tool in research and/or clinical practice. The IRI may assist in (1) early detection of disease transition in chronic diseases and (2) monitoring personal disease progression. It facilitates the detection of small deviations in clinical measurements, either using standard clinical biochemistry test results or omics data. With adequate data infrastructure, the IRIS workflow can be integrated into clinical practice by embedding personalised biomarker baselines into AI-enabled decision support systems Such integration provides a complementary layer that enhances the sensitivity and specificity of clinical alerts and risk stratification. Ultimately, this approach has the potential to transform personalised disease diagnosis, management, and patient outcomes.</p>

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

Reference intervals reimagined with IRIS for earlier detection and better disease monitoring

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年5月22日

查看全文
  • Murih Pusparum,
  • Wendy P. J. den Elzen,
  • Olivier Thas,
  • Gökhan Ertaylan

摘要

Reference intervals (RIs) are the universal established methodology for interpreting numerical clinical data by comparing individual test results to population-based benchmarks. The quality of these RIs significantly influences individual-level decision-making. This study aims to answer the question on how we can effectively leverage an individual’s personal data in conjunction with that of their peers to compute individual reference intervals (IRIs) for key clinical parameters relevant to disease detection and progression. We describe the IRIS workflow that includes prior data processing and data quality check procedures. The computation of IRI involves the test results of multiple “healthy” data points from the same subject(s) and also from the peers. The model adjusts for covariates like sex and age, enhancing accuracy. The workflow demonstrated the potential utility of IRI in clinical and omics data from two longitudinal studies. For healthy populations, IRIs showed diagnostic value in chronic diseases, while in diseased cohorts, they enabled effective disease monitoring. An integrated application IRIS has been developed, incorporating all described steps in an easy-to-use tool in research and/or clinical practice. The IRI may assist in (1) early detection of disease transition in chronic diseases and (2) monitoring personal disease progression. It facilitates the detection of small deviations in clinical measurements, either using standard clinical biochemistry test results or omics data. With adequate data infrastructure, the IRIS workflow can be integrated into clinical practice by embedding personalised biomarker baselines into AI-enabled decision support systems Such integration provides a complementary layer that enhances the sensitivity and specificity of clinical alerts and risk stratification. Ultimately, this approach has the potential to transform personalised disease diagnosis, management, and patient outcomes.