Artificial intelligence (AI) and machine learning (ML) have the potential to revolutionize workflows in the intensive care unit (ICU).
According to Dr Nils Schweingruber of the University Medical Center Hamburg-Eppendorf in Hamburg, Germany, several challenges must be met before ML models can be applied in the ICU: algorithms must be able to integrate complex and heterogenous data in real time, and the output must be easily interpretable.