Photo: Janette Pesu
In his doctoral dissertation, Antti Kallonen studied how artificial intelligence can predict the deterioration of patients in intensive care and emergency care settings more effectively than current methods. The most significant result of the dissertation concerns the early detection of sepsis in preterm infants: a neural network analyzing biosignals from patient monitors detected the condition typically about two days before clinical suspicion arose. At the same time, the AI identified new early markers in the biosignals that have not previously been used in recognizing severe infection. The results suggest that earlier detection of sepsis using artificial intelligence may enable earlier initiation of antibiotic treatment, which could reduce mortality and morbidity caused by sepsis.
The doctoral dissertation of MSc. Antti Kallonen in the field of biomedical engineering titled Artificial Intelligence in Critical Care Medicine: From Early Warning Scores to Multimodal Deep Learning will be publicly examined at the Faculty of Medicine and Health Technology at Tampere University on 27 February 2026.
The opponent will be Adjunct Professor Jukka Kortelainen from the University of Oulu. The custos will be Adjunct Professor Alpo Värri from the Faculty of Medicine and Health Technology, Tampere University.