Fig. 1From: Machine learning identifies ICU outcome predictors in a multicenter COVID-19 cohortDescriptive data of patients included into the study population. (n = 596 retrospective cohort and n = 443 prospective cohort). A Distribution Age B Horovitz quotient at admission C Murray lung injury score and SOFA score without GCS at admission D Survival rates E Interaction of Murray long injury score and admission status F Laboratory values. Grey indicates patients that did not survive ICU therapy, orange indicates patients that did survive ICU therapyBack to article page