Predictive Analysis of Hemodynamic Markers for Red Blood Cell Transfusion Outcomes in Anemia Pediatric Patients
Approximately half of pediatric intensive care unit (PICU) patients receive red-cell transfusions. However, those in a stable condition may endure the reduced oxygen delivery accompanying moderate anemia. A previous study explored physiological markers and assessed the ideal hemoglobin (Hb) threshold for red blood cell (RBC) transfusions in critically ill children. Furthermore, the study identified a set of 33 static and dynamic physiological parameters that may important in assessing the need of RBC transfusion.
Our objective is to harness advanced machine-learning techniques to sift through the 33 identified physiological markers. We aim to discern which markers most indicate anemia intolerance and, thus, are most influential in transfusion decision-making. Our proposed work is expected to enable clinicians make more informed, data-driven decisions that could enhance patient outcomes in pediatric critical care settings; this will be achieved with personalized medicine, where decisions are tailored to the individual patient’s physiological profile rather than a one-size-fits-all threshold based solely on Hb values.
Our study will be conducted at the PICU of CHU Sainte-Justine University Hospital (CHUSJ) and will include a cohort of children aged 0–18. We aim to evaluate the efficacy of red blood cell transfusion protocols in the pediatric intensive care setting. We plan to analyze the data that has already been acquired during previous studies and is available in the hospital database for research.
Required knowledge
Le candidat doit démontrer :
• Une connaissance des algorithmes d’intelligence artificielle
• Une excellente motivation
• Un bon dossier académique et/ou en recherche
• Capacité à travailler en équipe et d’une façon autonome
• Capacité à bien communiquer par écrit
• Excellente connaissance de programmation