The researchers from the Medical University of Bialystok (Michał Burdukiewicz, Jarosław Chilimoniuk, Krystyna Grzesiak, Adam Krętowski, Michał Ciborowski) published the review article „ML-based clinical decision support models based on metabolomics data” in TrAC Trends in Analytical Chemistry.
The authors discuss how modern medicine uses metabolomic data in clinical decision support systems (CDSS). CDSSs are widely used in medicine to facilitate critical decisions affecting patient health. CDSS includes a range of tools to help in daily clinical work, such as diagnostics. Metabolomics is a comprehensive study of metabolites, the products of cellular processes, that allow for a comprehensive description of health. However, metabolomic data is so complex that its comprehensive analysis requires machine learning methods.
In their paper, the authors outline the potential of building a CDSS based on metabolomic data. The authors also describe the risks associated with the lack of interpretability of such models and the lack of adequate verification of decision rules proposed by artificial intelligence.
The research was supported by Polish Ministry of Science and Higher Education within the project “Excellence Initiative–Research University” and Medical University of Białystok grant (B.SUB.24.548 to M.C.).
Link to the article: https://doi.org/10.1016/j.trac.2024.117819