May 15, 2026
Our research team at the University of Szeged, led by Dr. Pál Pásztor, has received an Innovation Award and Proof of Concept support for the DoseLearn project.
DoseLearn is developing a clinical protocol and a machine learning-based decision support system to help doctors plan how children in intensive care can be weaned off long-term sedatives and painkillers in a safer and more personalized way. This is a major challenge in pediatric intensive care: children can develop tolerance to these drugs within just a few days, which may lead to withdrawal symptoms or delirium.
In addition to prediction, the planned application will also support the continuous monitoring of the different phases of sedation. It aims to bring several key clinical tasks into one integrated system, including recording scoring results, calculating medication, and following patient history. According to clinicians, there is currently no dedicated bedside charting or other software solution available to them that can manage scoring, medication planning, and clinical history in such an integrated way.
The project combines eight years of clinical experience, international recommendations, scoring systems, and five years of patient data gathered by the clinic. Based on early results, the model may estimate the ideal weaning schedule more accurately and significantly reduce both the weaning period and the time spent in intensive care.
The software support for the project is carried out by the Department of Software Engineering, whose staff contribute to the development of the application and provide expertise related to the software implementation of AI-based prediction.
The next step is to build the application and validate it in multiple pediatric intensive care units, creating a broader research platform based on anonymous clinical data. This could make DoseLearn important not only for Szeged, but also for international pediatric intensive care practice.
Read more about the project on SZTE’s website!