Katharina Buczek
“Mathematically, this is something quite different. There has been little research in this regard in the medical field.”
Essentially, the AI makes its own decisions. And if the patient is doing well, the computer will be “rewarded.” On the other hand, if the patient’s condition declines or the patient dies, the computer will be “punished.”
So, the AI is tasked with reaching the highest possible “virtual reward” by completing successful actions.
And through this strategy, the AI uses various medical data to figure out a strategy that can achieve a high probability of success.
The team decided to center this study effort around sepsis treatment since sepsis is one of the most frequent causes of death in intensive care.
The condition is also extremely difficult for healthcare professionals to detect early, despite treatment being critical for patient survival.
Additionally, there have not been many medical breakthroughs in this area– meaning that the identification of new treatments is even more crucial.
“For this reason, it is particularly interesting to investigate the extent to which artificial intelligence can contribute to improving medical care here. Using machine learning models and other AI technologies is an opportunity to improve the diagnosis and treatment of sepsis, ultimately increasing the chances of patient survival,” Heitzinger detailed.
According to the study’s analysis, AI capabilities were even found to outperform humans. The researchers showed that by using AI, the cure rate for sepsis patients in regards to 90-day mortality was increased by 3%– totaling an approximate 88% cure rate.
Now, this success does not mean that AI will be solely responsible for making medical decisions in the ICU.
However, the technology could be extremely helpful for medical staff consultations, allowing doctors to compare their own assessments and suggestions against the AI’s findings.
