Impact of analytical bias on machine learning models for sepsis prediction using laboratory data.

Journal: Clinical chemistry and laboratory medicine
Published Date:

Abstract

OBJECTIVES: Machine learning (ML) models, using laboratory data, support early sepsis prediction. However, analytical bias in laboratory measurements can compromise their performance and validity in real-world settings. We aimed to evaluate how analytically acceptable bias may affect the validity and generalizability of ML models trained on laboratory data.

Authors

  • Meryem Rümeysa Yeşil
    Department of Medical Biochemistry, University of Health Sciences, Bursa Yuksek Ihtisas Training and Research Hospital, Bursa, Turkey.
  • Ilaria Talli
    Department of Medicine (DIMED), University of Padova, Padova, Italy; Laboratory Medicine Unit, University-Hospital, Padova, Italy; QI.LAB.MED, Spin-off of the University of Padova, Italy.
  • Michela Pelloso
    Laboratory Medicine Unit, University-Hospital, Padova, Italy.
  • Chiara Cosma
    Department of Medicine (DIMED), University of Padova, Padova, Italy; QI.LAB.MED, Spin-off of the University of Padova, Italy.
  • Elisa Pangrazzi
    Department of Medicine (DIMED), University of Padova, Padova, Italy; QI.LAB.MED, Spin-off of the University of Padova, Italy.
  • Mario Plebani
    Department of Laboratory Medicine, University-Hospital of Padova, Padova, Italy.
  • Yasemin Üstündağ
    Department of Medical Biochemistry, University of Health Sciences, Bursa Yuksek Ihtisas Training and Research Hospital, Bursa, Turkey.
  • Andrea Padoan
    Department of Laboratory Medicine, University-Hospital of Padova, via Giustiniani 2, Padova 35128, Italy.

Keywords

No keywords available for this article.