Related Experiment Video
Updated: Aug 27, 2025
Performing Data Mining And Integrative Analysis Of Biomarker in Breast Cancer Using Multiple Publicly Accessible Databases
Published on: May 17, 2019
Inductive database to support iterative data mining: Application to biomarker analysis on patient data in the
Emmanuel Bresso1, Joao-Pedro Ferreira2, Nicolas Girerd2
1Université de Lorraine, CNRS, Inria Nancy G.E., LORIA, UMR 7503, Vandoeuvre-lès-Nancy, France; Université de Lorraine, Centre d'Investigations Cliniques Plurithématique 1433, INSERM 1116, CHRU de Nancy, France.
This study introduces an Inductive Clinical DataBase (ICDB) to integrate machine learning into health systems for better knowledge discovery. The ICDB approach successfully identified predictive biomarkers for heart failure, advancing the development of next-generation knowledge discovery environments.
Area of Science:
- Biomedical informatics
- Machine learning in healthcare
- Data mining
Background:
- Machine learning is crucial for biomedical studies but not fully integrated into Learning Health Systems.
- The Knowledge Discovery from Data (KDD) process requires enhanced integration for clinical applications.
Purpose of the Study:
- To propose an extension of the KDD process model using an inductive database.
- To introduce a generic model for an Inductive Clinical DataBase (ICDB) capable of storing patient data and learned models.
- To demonstrate the utility of the ICDB approach in identifying predictive biomarkers for heart failure.
Main Methods:
- Developed a novel Inductive Clinical DataBase (ICDB) model.
- Applied the ICDB to patient data for heart failure research.
- Evaluated two KDD scenarios: local-to-global and trans-cohort alignment.
Main Results:
- The ICDB approach successfully identified biomarker combinations predictive of heart fibrosis phenotype.
- Hypotheses regarding underlying mechanisms of heart failure were generated.
- Demonstrated the feasibility of integrating patient data and learned models within a clinical database.
Conclusions:
- The ICDB model offers a promising framework for advancing Knowledge Discovery in healthcare.
- This proof of concept paves the way for next-generation Knowledge Discovery Environments (KDE).
- The approach facilitates the discovery of novel insights from clinical data for disease mechanisms.
