Related Experiment Video
Updated: Jun 27, 2026
05:35
A Point-of-Care Method with Integrated Decision Support Tool to Estimate Anemia at Population Level
Published on: January 19, 2024
Predicting Iron Deficiencies Using Routine Complete Blood Cell Count Parameters: A Machine Learning Approach and
Davide Negrini1, Laura Pighi1, Simone Mignolli1
1Section of Clinical Biochemistry, University of Verona, 37134 Verona, Italy.
Journal of Clinical Medicine
|June 26, 2026
Summary
Machine learning models using complete blood count (CBC) can predict low iron levels. This approach aids in identifying individuals needing further iron status assessment, improving diagnostic efficiency.
Area of Science:
- Hematology
- Medical Informatics
- Machine Learning
Background:
- Iron deficiency is a widespread health issue requiring specific diagnostic tests.
- Routine complete blood count (CBC) parameters are readily available.
- Targeted laboratory testing can optimize resource utilization.
Purpose of the Study:
- To evaluate machine learning models using CBC parameters for predicting low ferritin and transferrin saturation.
- To assess the potential of CBC-based models in guiding iron status testing.
Main Methods:
- Retrospective analysis of 32,437 outpatient records with CBC and iron metabolism tests.
- Trained multiple supervised machine learning models (e.g., random forest, XGBoost) on CBC indices.
- Validated model performance using area under the curve (AUC), sensitivity, and specificity.
Main Results:
- All models demonstrated predictive capability for low iron markers using CBC data alone.
- Random forest and XGBoost achieved the highest performance (AUC 0.80-0.96).
- Key predictors included mean corpuscular volume (MCV), mean corpuscular hemoglobin (MCH), and red blood cell distribution width (RDW).
Conclusions:
- Machine learning models utilizing CBC parameters can effectively identify individuals with potential iron deficiency.
- These models support targeted iron status assessment, enhancing diagnostic strategies.
