Georgia Charkoftaki, the study’s lead author and an associate research scientist at Yale, said, “Our AI-powered patient triage platform is different from traditional Covid-19 AI prediction algorithms. With the help of artificial intelligence (AI), researchers have created a cutting-edge patient triage platform that, according to their claims, can gauge the severity of a patient’s illness and the length of their hospital stay during a viral outbreak.
According to the severity, prognosis, and availability of resources, patient care is prioritized based on their sickness. The platform, which was recently detailed in the journal Human Genomics, makes use of data from metabolomics, the study of tiny chemicals connected to cell metabolism, as well as machine learning, a type of AI.
According to the researchers, the innovation is meant to enhance patient management and enable healthcare professionals in allocating resources more effectively amid severe virus epidemics that can swiftly overwhelm local healthcare systems.
According to senior study author Vasilis Vasiliou, a professor at Yale University in the US, “being able to predict which patients can be sent home and those who may need to intensive care unit admission is critical for health officials seeking to optimise patient health outcomes and use hospital resources most efficiently during an outbreak.”
It bases its predictions on the integration of routine clinical data, patient comorbidity data, and untargeted plasma metabolomics data. Based on clinical information and metabolic profiles gathered from COVID-19 patients hospitalized with the disease, the researchers developed a model of COVID-19 severity and prediction of hospitalization using machine learning.
The model allowed us to discover a group of distinct clinical and metabolic biomarkers that were highly suggestive of disease progression and allowed the prediction of patient treatment requirements relatively quickly after hospitalization, according to the study’s authors.